rare earth elements - sgtk - news · rare earth elements 771 table 2. estimated annual world mine...

Rare Earth ElementsStephen B. Castor and James B. Hedrick

INTRODUCTIONThe rare earth elements (REEs), which include the 15 lanthanideelements (Z = 57 through 71) and yttrium (Z = 39), are so calledbecause most of them were originally isolated in the 18th and 19thcenturies as oxides from rare minerals. Because of their reactivity,the REEs were found to be difficult to refine to pure metal. Further-more, efficient separation processes were not developed until the20th century because of REEs’ chemical similarity. A statementattributed to Sir William Crookes, a noted English scientist, betraysthe frustrations of late 19th-century chemists with this group of ele-ments (Emsley 2001):

The rare earth elements perplex us in our researches, baf-fle us in our speculations, and haunt us in our verydreams. They stretch like an unknown sea before us,mocking, mystifying and murmuring strange revelationsand possibilities.

All of the REEs were finally identified in the 20th century.Promethium, the rarest, was not identified until 1945, and pure lute-tium metal was not refined until 1953 (Emsley 2001). Commercialmarkets for most of the REEs have arisen in only the past 50 years.

Most REEs are not as uncommon in nature as the nameimplies. Cerium, the most abundant REE (Table 1), comprises moreof the earth’s crust than copper or lead. Many REEs are more com-mon than tin and molybdenum, and all but promethium are morecommon than silver or mercury (Taylor and McClennan 1985).

Promethium, best known as an artificial element, occurs invery minute quantities in natural materials because it has no stableor long-lived isotopes. Lanthanide elements with low atomic num-bers are generally more abundant in the earth’s crust than thosewith high atomic numbers. Those with even atomic numbers aretwo to seven times more abundant than adjacent lanthanides withodd atomic numbers (Table 1).

The lanthanide elements traditionally have been divided intotwo groups: the light rare earth elements (LREEs)—lanthanumthrough europium (Z = 57 through 63); and the heavy rare earth ele-ments (HREEs)—gadolinium through lutetium (Z = 64 through71). Although yttrium is the lightest REE, it is usually grouped withthe HREEs to which it is chemically and physically similar.

The REEs are lithophile elements (elements enriched in theearth’s crust) that invariably occur together naturally because all aretrivalent (except for Ce+4 and Eu+2 in some environments) and have

76

similar ionic radii. An increase in atomic number in the lanthanidegroup is not accompanied by change in valence, and the lanthanideelements all inhabit the same cell in most versions of the periodictable. The similar radii and oxidation states of the REEs allow forliberal substitution of the REEs for each other into various crystallattices. This substitution accounts for their wide dispersion in theearth’s crust and the characteristic multiple occurrences of REEswithin a single mineral. The chemical and physical differences thatexist within the REEs group are caused by small differences inionic radius and generally result in segregation of REEs into depos-its enriched in either light lanthanides or heavy lanthanides plusyttrium.

The relative abundance of individual lanthanide elements hasbeen found useful in the understanding of magmatic processes andnatural aqueous systems. Comparisons are generally made using a

Table 1. REEs, atomic numbers, and abundances

Element SymbolAtomicNumber

Upper Crust Abundance,

ppm*

Chondrite Abundance,

ppm†

Yttrium Y 39 22 na‡

Lanthanum La 57 30 0.34

Cerium Ce 58 64 0.91

Praseodymium Pr 59 7.1 0.121

Neodymium Nd 60 26 0.64

Promethium Pm 61 na na

Samarium Sm 62 4.5 0.195

Europium Eu 63 0.88 0.073

Gadolinium Gd 64 3.8 0.26

Terbium Tb 65 0.64 0.047

Dysprosium Dy 66 3.5 0.30

Holmium Ho 67 0.80 0.078

Erbium Er 68 2.3 0.20

Thulium Tm 69 0.33 0.032

Ytterbium Yb 70 2.2 0.22

Lutetium Lu 71 0.32 0.034

* Source: Taylor and McClennan 1985† Source: Wakita, Rey, and Schmitt 1971.‡ na = not available.

9

770 Industrial Minerals and Rocks

logarithmic plot of lanthanide abundances normalized to abun-dances in chondritic (stony) meteorites. The use of this methodeliminates the abundance variation between lanthanides of odd andeven atomic number, and allows determination of the extent of frac-tionation between the lanthanides, because such fractionation is notconsidered to have taken place during chondrite formation. Themethod also is useful because chondrites are thought to be compo-sitionally similar to the original earth’s mantle. Europium (Eu)anomalies (positive or negative departures of europium from chon-drite-normalized plots) have been found to be particularly effectivefor petrogenetic modeling. In addition, REE isotopes, particularlyof neodymium and samarium, have found use in petrogenetic mod-eling and geochronology.

HISTORY OF REE PRODUCTIONREEs were originally produced in minor amounts from smalldeposits in granitic pegmatite, the geologic environment in whichthey were first discovered. During the second half of the 19th cen-tury and the first half of the 20th century, REEs came mainly fromplacer deposits, particularly those of the southeastern UnitedStates. With the exception of the most abundant lanthanide ele-ments (cerium, lanthanum, and neodymium), individual REEswere not commercially available until the 1940s. Between 1965and 1985, most of the world’s REEs came from Mountain Pass,California. Heavy mineral sands from placers in many parts of theworld, however, were also sources of by-product REE minerals,and Australia was a major producer from such sources until theearly 1990s. Until recently, Russia also was an important REE pro-ducer from a hard rock source. During the 1980s, China emerged asa major producer of REE raw materials, while the Australian andAmerican market shares decreased dramatically (Figure 1). Since1998, more than 80% of the world’s REE raw materials have comefrom China, and most of this production is from the Bayan Obodeposit in Inner Mongolia. Table 2 gives recent annual productionfigures by country, and Figure 2 shows locations of currently andrecently productive REE mines.

MINERALS THAT CONTAIN REESAlthough REEs comprise significant amounts of many minerals,almost all production has come from less than 10 minerals. Table 3lists those that have yielded REEs commercially or have potentialfor production in the future. Extraction from a potentially economicREE resource is strongly dependant on its REE mineralogy. In thepast, producing deposits were limited to those containing REE-bearing minerals that are relatively easy to concentrate because ofcoarse grain size or other attributes. Minerals that are easily brokendown, such as the carbonate bastnasite, are more desirable thanthose that are difficult to dissociate, such as the silicate allanite.Placer monazite, once an important source of REEs, has beenlargely abandoned because of its high thorium content. Recently,REEs absorbed on clay minerals in laterite have become importantsources of REEs in China. For more information on REE-bearingminerals, see works by Mariano (1989a); and Jones, Wall, and Wil-liams (1996).

GEOLOGY OF REE DEPOSITSIron-REE DepositsSome iron deposits contain REE resources, and such deposits havebeen exploited in only one area—Bayan Obo, China. These depos-its constitute the largest known REE resource in the world (Table 4)and are now the most important REE source in the world (Haxel,Hedrick, and Orris 2002).

Iron-LREE-niobium deposits at Bayan Obo in Inner Mongolia,China, were discovered by Russian geologists in 1927 (Argall 1980)when Inner Mongolia was under the control of the former U.S.S.R.REEs are recovered from iron-REE-niobium (Fe-REE-Nb) ore bod-ies that have been mined from more than 20 sites since miningstarted in 1957 (Drew, Meng, and Sun 1990). The two largest are theMain and East ore bodies (Figure 3), each of which include iron-REE resources with more than 1,000 m of strike length and average5.41% and 5.18% rare earth oxides (REOs), respectively (Yuan et al.1992). Total reserves have been reported as at least 1.5 billion t ofiron (average grade 35%), at least 48 Mt of REOs (average grade6%), and about 1 Mt of niobium (average grade 0.13%) (Drew,Meng, and Sun 1990). The REE ore consists of three major types:REE-iron ore, the most important type; REE ore in silicate rock;and REE ore in dolomite (Yuan et al. 1992). Massive REE-iron oreoccurs in 100-m-thick lenses with local banded and breccia struc-tures. Banded and streaky fluorite-rich REE-iron ore has the high-est REE content, locally more than 10% REOs. Major oxidechemistries of several Bayan Obo ore types are shown in Table 5.REEs are mainly in bastnasite and monazite, but at least 20 otherREE-bearing minerals have been identified (Yuan et al. 1992).Bayan Obo ore has extreme LREE enrichment with no europiumanomalies (Figure 4) and low strontium (87Sr/86Sr) (Nakai et al.1989; Philpotts et al. 1989). Alkali-rich alteration (fenitization),predominated by sodic amphibole and potash feldspar, is associatedwith the REE mineralization (Drew, Meng, and Sun 1990).

The Bayan Obo ore is hosted by dolomite of the Bayan OboGroup, a Middle Proterozoic clastic and carbonate sedimentarysequence (Qiu, Wang, and Zhao 1983; Chao et al. 1997) that occursin an 18-km-long syncline (Drew, Meng, and Sun 1990). TheBayan Obo Group was deposited unconformably on 2.35-Ga mig-matites, and, along with Carboniferous volcanics, was deformedduring a Permian continent-to-continent collision event dominatedby folding and thrusting (Drew, Meng, and Sun 1990). Intrusion oflarge amounts of Permian granitoid rocks also resulted from thiscollision.

The age of the REE mineralization is a matter of debate. Ithas been dated at 1600 Ma (Yuan et al. 1992) and 550 to 400 Ma(Chao et al. 1997). Most researchers agree that Fe-REE-Nb miner-alization at Bayan Obo took place over a long period of time.

Figure 1. Production of REEs by China, the United States, Australia, and other countries, 1983–2003

China

0

10

20

30

40

50

60

70

80

90

100

1983 1988 1993 1998 2003

Prod

uctio

n, k

t

Other*Australia

United States

*“Other” includes India, Brazil, Kyrgyzstan, Sri Lanka, Russia, Malaysia, and Thailand.

Rare Earth Elements 771

Table 2. Estimated annual world mine production of REEs, by country

Country

1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003*

Metric Tons of REO Equivalent

Australia 8,328 10,304 7,047 7,150 3,850 1,650 110 0 0 0 0

Brazil 2,891 2,174 2,383 1,377 719 270 103 0 0 0 0

China na† 8,500 15,100 25,220 16,150 22,100 48,000 53,000 70,000 80,600 90,000

India 2,200 2,200 2,200 2,365 2,200 2,500 2,750 2,750 2,700 2,700 2,700

Kyrgyzstan na na na 696 721 0 na na 6,115 3,800 na

Malaysia 601 3,869 1,618 1,700 1,093 224 452 422 631 281 450

Mozambique 2 2 0 0 0 0 0 0 0 0 0

South Africa 0 0 660 660 237 237 0 0 0 0 0

Sri Lanka 165 110 110 110 110 110 110 110 120 0 0

Thailand 164 459 270 368 229 127 0 7 0 0 0

Russia na na na 7,626 6,138 4,468 2,000 2,000 2,000 2,000 2,000

United States 17,083 13,428 11,100 20,787 16,465 17,754 22,200 10,000 5,000 5,000 0

Zaire 6 0 53 96 66 11 5 0 0 0 0

Total 31,439 41,047 40,541 68,155 47,978 49,449 75,730 68,288 86,566 94,381 95,150

Courtesy of USGS.

* Estimated. Some data have been added or modified using unpublished data from USGS files.† na = not available.

Figure 2. Locations of the world’s rare earth mines

Currently operating minesRecently active and potential new sources

17

7 18

19

20

8

9

5 10

11

6

34

12

1213

14 1516

Deposit TypeLocation NameLocation

No.

Monazite by-product, coastal placersMonazite by-product, coastal placersMonazite by-product, coastal placersLateritized carbonatiteAltered alkaline complexMonazite by-product, coastal placersUraniferous conglomerateMonazite by-product, placerMonazite by-product, coastal placersMonazite-apatite vein

Orissa, IndiaEneabba, AustraliaCapel and Yoganup, AustraliaMount Weld, AustraliaDubbo, AustraliaNorth Stradboke Island, AustraliaElliot Lake, CanadaGreen Cove Springs, USACamaratuba, BrazilSteenkampskraal, South Africa

11121314151617181920

Deposit TypeLocation NameLocation

No.

Fe-REE-Nb depositBastnasite-barite veinsBastnasite-barite veinsLateritic clayMonazite by-product, coastal placersXenotime by-product, tin placersBastnasite-barite carbonatiteLoparite in peralkaline complexPolymetallic depositMonazite by-product, coastal placers

Bayan Obo, ChinaWeishan, ChinaMaoniuping, ChinaXunwu and Longnan, ChinaChavara, IndiaPerak, MalaysiaMountain Pass, USALovozero, RussiaAktyus, KyrgyzstanNorthern Sri Lanka

123456789

10


Recent research points to ore formation during hydrothermalreplacement of sedimentary dolomite, possibly related to one of thefollowing:

• Alkaline or carbonatite magmatism (Drew, Meng, and Sun1990; Campbell and Henderson 1997; Smith and Henderson2000; Smith, Henderson, and Campbell 2000)

• Original emplacement of both the host dolomite and ore asigneous carbonatite with subsequent tectonic deformation (LeBas, Spiro, and Yang 1997)

• Formation of host and ore by marine exhalation of alkalic orcarbonatite-related fluids Whatever its origin, the Bayan Obo deposits now dominate

worldwide REE production and probably will do so for many yearsto come.

Table 3. Minerals that contain REEs and occur in economic or potentially economic deposits

Mineral Formula* REO

wt %† ‡

Aeschynite (Ln,Ca,Fe,Th)(Ti,Nb)2(O,OH)6 36

Allanite (orthite) (Ca,Ln)2(Al,Fe)3(SiO4)3(OH) 30

Anatase TiO2 3

Ancylite SrLn(CO3)2(OH)•H2O 46

Apatite Ca5(PO4)3(F,Cl,OH) 19

Bastnasite LnCO3F 76

Brannerite (U,Ca,Ln)(Ti,Fe)2O6 6

Britholite (Ln,Ca)5(SiO4,PO4)3(OH,F) 62

Cerianite (Ce,Th)O2 81§

Cheralite (Ln,Ca,Th)(P,Si)O4 5

Churchite YPO4•2H2O 44‡

Eudialyte Na15Ca6(Fe,Mn)3Zr3(Si,Nb)Si25O73(OH,Cl,H2O)5

10

Euxenite (Ln,Ca,U,Th)(Nb,Ta,Ti)2O6 <40§

Fergusonite Ln(Nb,Ti)O4 47

Florencite LnAl3(PO4)2(OH)6 32§

Gadolinite LnFeBe2Si2O10 52

Huanghoite BaLn(CO3)2F 38

Hydroxylbastnasite LnCO3(OH,F) 75

Kainosite Ca2(Y,Ln)2Si4O12CO3•H2O 38

Loparite (Ln,Na,Ca)(Ti,Nb)O3 36

Monazite (Ln,Th)PO4 71

Mosandrite (Ca,Na,Ln)12(Ti,Zr)2Si7O31H6F4 <65§

Parisite CaLn2(CO3)3F2 64

Samarskite (Ln,U,Fe)3(Nb,Ta,Ti)5O16 12

Synchisite CaLn(CO3)2F 51

Thalenite Y3Si3O10(OH) 63§

Xenotime YPO4 61§

Yttrotantalite (Y,U,Fe)(Ta,Nb)O4 <24§

* Source for mineral formulas: Mandarino 1999, with Ln = lanthanide elements.

† Sources for REO content: Frondel 1958; Overstreet 1967; Anon. 1980; Kapustin 1980; Mazzi and Munno 1983; Mariano 1989a.

‡ Where more than one analysis is available, the analysis with the highest REO content is reported (e.g., REO for monazite from the Mountain Pass carbonatite is reported; monazite from pegmatites and metamorphic rocks generally has lower REO).

§ Stoichiometric calculation of REO content.

The Pea Ridge iron deposit in Missouri contains an unex-ploited high-grade REE resource of unknown but probably smallsize in breccia pipes associated with 1.48-Ga granite and syenite(Kisvarsanyi, Sims, and Kisvarsanyi 1989). Bulk samples of REE-rich breccia, which contain monazite and xenotime, average about12% REEs (Nuelle et al. 1992). Although the Pea Ridge resource isLREE-dominated, it also contains significant HREEs.

Mill tailings from processing magnetite deposits in Precam-brian gabbro and syenite at Mineville, New York, are an REEresource of unknown size. They contain apatite that averages morethan 11% REOs, including about 2% Y2O3 (McKeown and Klemic1956).

The important magnetite and hematite deposits at Kiruna,Sweden, average 0.7% and 0.5% REOs, respectively, and otherSwedish iron deposits contain similar amounts of REOs (Parak1973). The REEs in these deposits are reportedly in apatite andmonazite.

The huge Middle Proterozoic Olympic Dam copper-uranium-silver-gold deposit, South Australia, may also be included in thisdeposit type because most of the ore is breccia that contains 40% to90% hematite along with quartz, sericite, fluorite, barite, sulfides,and REE minerals (Oreskes and Einaudi 1990). Olympic Dam is apotential source of by-product REEs, but typical content is about0.5% total REEs, and there are no plans for REE production fromthe deposit. Extremely fine-grained monazite and bastnasite are themost abundant REE minerals, and the LREE/HREE ratio is high inOlympic Dam ore.

Carbonatite DepositsMany carbonatite intrusions are enriched in REEs; Orris andGrauch (2002) list more than 100 carbonatite occurrences that con-tain REE minerals. Brief descriptions of REE mineral occurrencesin carbonatites worldwide may be found in works by Mariano(1989a), Woolley (1987), and Wall and Mariano (1996).

Despite the abundance of REE-bearing carbonatites, REEs havebeen produced from only one, the Mountain Pass deposit in Califor-nia. Although the Mountain Pass mine and plant did not operate in2003, REE products were shipped from stockpile (Hedrick 2004a).Beginning in 1954, the Mountain Pass deposit was mined exclusivelyfor REEs by Molycorp, Inc. (originally Molybdenum Corp. of Amer-ica and now a subsidiary of Unocal). In 1987, proven and probablereserves totaled 29 Mt with an average grade of 8.9% REOs based ona 5% cutoff grade (Castor 1991). Current reserves are in excess of20 Mt of ore at similar grade (Castor and Nason 2004). The ore typi-cally contains 10% to 15% bastnasite, 65% calcite or dolomite, and20% to 25% barite. Other gangue minerals, such as strontianite andtalc, are present in significant amounts locally. Galena is locallyabundant, but other sulfide minerals are rare. Bastnasite is the onlymineral processed, although nine other REE minerals occur at Moun-tain Pass (Castor and Nason 2004). Mountain Pass ore has highLREEs and HREEs (Figure 4). The major oxide chemistry of Moun-tain Pass REE ore types is shown in Table 5.

The Mountain Pass carbonatite, which has been dated at about1.4 Ga (DeWitt, Kwak, and Zartman 1987), is a moderately dip-ping, tabular intrusion (Figure 5) into granulite-grade gneiss. It isassociated with ultrapotassic alkaline plutons of similar age, size,and orientation, as well as with abundant carbonatite and alkalinedikes, and is in a narrow north-trending zone of ultrapotassic alka-line igneous rocks at least 130 km long (Castor 1991; Castor andNason 2004).

Elsewhere in the western United States, carbonatites havebeen investigated as possible sources of REEs, but none rival theMountain Pass deposit as a REE resource. Similar to the Mountain


Table 4. Size and grade of some productive and potentially productive REE deposits

Deposit or District Country Size, t REOGrade,% REO Deposit Type

Recent Production Reference

Bayan Obo China 48,000,000 6 Iron-rich Yes Drew, Meng, and Sun 1990

Araxá Brazil 8,100,000 1.8 Carbonatite laterite No Morteani and Preinfalk 1996

Mountain Pass United States 1,800,000 8.9 Carbonatite Yes Castor and Nason 2004

Mount Weld Australia 1,700,000 11.2 Carbonatite laterite No Mariano 1989b

Dubbo Australia 700,000 0.86 Trachyte No Australian Zirconia Ltd. 2000

Mrima Hill Kenya 300,000 5 Carbonatite laterite No Orris and Grauch 2002

Nolan's Bore Australia 150,000 4 Vein No Anon. 2003

Xunwu and Longnan China Unknown 0.05–0.2 Laterite Yes Wu, Yuan, and Bai 1996

Lovozero Russia Unknown 0.01 Peralkaline syenite Yes Hedrick, Sinha, and Kosynkin 1997

Maoniuping China Unknown 2 Vein Yes Wu, Yuan, and Bai 1996

Weishan China Unknown 1.6 Vein Yes Wu, Yuan, and Bai 1996

Aktyus Kyrgyzstan Unknown 0.25 Unknown Yes Geological Survey of Kyrgyzstan 2004

Eneabba Australia Unknown 0.001 Placer Yes Griffiths 1984

Pass carbonatite, most have high LREE/HREE ratios. The largeCambrian Iron Hill carbonatite in the Powderhorn District of Colo-rado has been considered a possible source of REEs, but Armbrust-macher (1980) reported a grade of about 0.5% REEs. Carbonatitedikes in the Powderhorn District that contain up to 3% REOs con-stitute a resource of unknown size (Olson and Hedlund 1981), anddikes in the nearby Wet Mountains constitute a small resource witha grade of about 2.5% REOs (Armbrustmacher 1988). Phanerozoiccarbonatite dikes that contain as much as 21% REOs (Crowley1960), mainly in monazite, allanite, and ancylite (Heinrich andLevinson 1961), are present in the North Fork Area on the Idaho–Montana border, but the size of this resource is probably smallbecause dikes with high REO contents are generally less than 1 mthick.

Carbonatite complexes in eastern Canada also contain REEmineralization. A britholite-bearing carbonatite dike in Mesozoiccarbonatite at Oka, Quebec, is an REE resource of unknown sizeand grade (Mariano 1989a). Bastnasite and monazite zones con-taining up to 4.5% LREEs have been identified in the Late Protero-zoic St-Honore Complex in Quebec, with one 80-m drill interceptaveraging more than 3% REOs (Vallee and Dubuc 1970).

Brazilian carbonatites contain significant REE resources buthave yielded only minor production to date. Niobium ore in theBarreiro Complex at Araxá, Minas Gerais, which comprises450+ Mt averaging 2.5% Nb2O5, contains about 4.4% REOs(Issa Filho, Lima, and Souza 1984).

African carbonatites also contain LREE-dominated depositsthat are largely unexploited. Kangankunde Hill, Malawi, is under-lain by carbonatite dikes in which the main REE phase is monazitethat is almost thorium-free (Garson 1966). The dikes are thought tocontain a resource of several hundred thousand tons of monazite atan average grade of more than 5% (Deans 1966). Mariano (1989a)suggested that monazite in Kangankunde Hill carbonatite is of pri-mary igneous origin like the bastnasite at Mountain Pass, but subse-quent work indicated that the monazite, as well as associatedbastnasite and florencite-goyazite, was formed from an originalREE phase during hydrothermal activity (Wall and Mariano 1996).Carbonatite dikes at Wigu Hill, Tanzania, contain up to 20% REOsas monazite and REE fluorocarbonate minerals of possible hydro-thermal origin (Deans 1966). The Palabora carbonatite in SouthAfrica, which occurs in a large Early Proterozoic potassic alkalinecomplex, has significant copper and apatite production and has beenevaluated as a source of REEs. Potential by-product from apatite

concentrates that contain 0.4% to 0.9% REOs has been estimated at4,250 tpy (Notholt, Highley, and Deans 1990).

Sill-like lenses of bastnasite-fluorite-barite rock as much as30 m thick with about 5% REEs are associated with carbonatite atKizilçaören, Turkey (Hatzl et al. 1990). The carbonatite occurs indikes associated with intrusive Tertiary phonolites and trachytesalong with alkaline pyroclastic rocks. The unexploited Kizil-çaören resource is dominantly LREEs (G. Morteani, personalcommunication).

Lateritic DepositsLateritic deposits that occur over low-grade primary sources, suchas carbonatites and syenites, have been studied as potential REEsources since the 1980s. Such deposits may constitute largeresources and some have high REE contents; however, only twodeposits in China have been exploited to date.

REEs have been produced in increasing quantities in recentyears from surficial clay deposits in southern China, and Orris andGrauch (2002) list 18 Chinese occurrences of such material. In1992, REEs from these deposits comprised 14% of Chinese pro-duction (Wu, Yuan, and Bai 1996), and this source has had astrong impact on yttrium supplies since 1988. The depositsreportedly form weathering crusts over granite (Ren 1985; Wu,

Adapted from Chao et al. 1997.Figure 3. Geologic map of the Main and East iron-REE-niobium ore bodies at Bayan Obo, China

Marble and quartzite

Disseminated Fe-REE ore

Banded REE-Fe-F ore

Massive Fe ore

Biotite schist Fe ore

Amphibole, pyroxene,or dolomitic Fe oreSlate and schist

500 m


Table 5. Major oxide chemistry of ore samples from Bayan Obo and Mountain Pass

Samples

Bayan Obo* Mountain Pass†

Mfe Ffe AeFe RiFe MDT 7B30-20 7B30-45A 7B30-28B 7B31-10 85-4 R-724

SiO2 4.81 2.18 23.86 10.79 8.74 na ‡ na na na 0.40 1.63

TiO2 0.27 0.62 0.56 0.55 0.28 1.00 0.15 0.23 0.08 0.01 0.01

Al2O3 0.22 0.66 1.55 0.83 0.74 0.06 0.25 0.17 0.26 0.01 0.01

Fe2O3§ 74.73 39.29 31.61 44.59 11.69 7.72 13.87 37.17 5.00 0.24 1.77

MnO 0.79 0.12 0.07 5.95 1.18 0.01 0.03 0.02 0.01 0.24 0.45

MgO 0.99 0.31 0.30 3.52 13.23 <0.03 0.07 0.25 0.12 0.04 6.41

CaO 8.78 26.26 11.59 16.15 27.09 27.98 27.98 3.92 8.67 21.30 11.73

SrO 0.36 3.90 5.67 1.15 0.25 0.05 0.10 0.07 0.09 14.15 2.49

BaO na na na na na 0.04 3.80 2.79 2.46 14.63 25.12

Na2O 0.25 0.25 4.06 0.62 0.12 na na na na 0.05 0.07

K2O 0.09 0.08 0.57 0.92 0.58 <0.24 <0.24 <0.24 <0.24 0.05 0.09

P2O5 0.94 2.71 2.85 1.16 1.47 na na na na 0.04 0.29

F 5.89 16.83 7.25 8.31 1.83 17.00 14.4 1.68 4.49 0.70 1.40

CO2 na na na na na na na na na 19.84 18.26

SO3 na na na na na na na na na 18.26 15.43

LOI 2.89 5.15 5.62 5.60 25.23 na na na na na na

RE2O3 2.73 9.49 7.72 3.24 3.98 19.32 8.67 10.89 42.75 9.89 13.18

Nb2O5 na na na na na 0.36 0.02 0.05 0.53 0.001 0.004

Total 103.74 107.85 103.28 103.38 96.41 73.54 69.34 57.24 64.46 99.85 98.34

Source: Yuan et al 1992; Chao et al. 1997; Castor and Nason 2004.* Bayan Obo:

Mfe = massive REE-Fe oreFfe = fluorite REE-Fe oreAeFe = aegerine REE-Fe oreRiFe = riebeckite REE-Fe oreMDT = magnetite-dolomite ore7B30-20 = high-REE, low-FE banded ore

7B30-45A = bastnasite-apatite-pyroxene-fluorite ore7B30-28B = high-Fe, REEs, and fluorite ore7B31-10 = high-REE, low-Fe, and low-F ore

† Mountain Pass: 85-4 = bastnasite-barite sovite, Sulphide Queen ore bodyR-724 = bastnasite-barite beforsite, Sulphide Queen ore body.

‡ na = not available.§ All Fe as Fe2O3.

Yuan, and Bai 1996). The ore, referred to as REE-bearing ionicabsorption clay, mostly comes from two sites in Jiangxi Prov-ince—Longnan and Xunwu, the former yielding HREE- andyttrium-rich material and the other, LREE-rich material(O’Driscoll 2003). Ore from Longnan has an HREE-dominated

Figure 4. Chondrite-normalized plot of REEs in ores from Bayan Obo, China, and Mountain Pass, United States (data sources: Castor 1986; Yuan et al. 1992)

1,000,000

100,000

10,000

1,000

100

10

1La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Sam

ple:

Cho

ndrit

e

NOTE: REE data not normalized to 100% prior to chondrite normalization.

Bayan Obo Fe-REE-aegerine ore

Bayan Obo Fe-REE-fluorite ore

Mountain Pass beforsite ore

Mountain Pass sovite ore

distribution pattern very similar to that of xenotime, whereas orefrom Xunwu is relatively enriched in lanthanum (Hedrick 1992;Figure 6). Both ores have relatively low cerium content, suggest-ing deposition from REE-bearing groundwater with depletedcerium that results from the element’s insolubility in the oxidized(Ce+4) state. The ore bodies are 3 to 10 m thick and occur mainlyin a wholly weathered zone composed of halloysite and kaolinitewith residual quartz and feldspar; grades are reported at 0.05% to0.2% REOs (Wu, Yuan, and Bai 1996). The deposits are consid-ered to be laterites and show similarities to a number of other lat-eritic deposits formed over alkaline igneous rocks and carbonatite(Morteani and Preinfalk 1996).

REE-enriched lateritic material overlying carbonatite ispresent in many sites worldwide, mostly in the tropics. Orris andGrauch (2002) list more than 50 such occurrences, but, other thanthe Chinese deposits, none have produced significant amounts ofREEs.

The Mount Weld REE deposit in southern Western Australiais in the laterized cap over a large carbonatite (Figure 7) dated at 2.0Ga (Duncan and Willett 1990). Although LREE/HREE ratios aregenerally high, the laterite is locally enriched in HREEs andyttrium. Published reserve figures are 15.4 Mt with 11.2% REOs +Y2O3 (Mariano 1989a), and 2.3 Mt with 18.0% REOs (Willis2002). Large volumes of ore contain an average of 0.33% Y2O3,and niobium and tantalum enrichment has also been reported(O’Driscoll 1988). Data from Lottermoser (1990) indicate thatunweathered carbonatite at depth contains 0.1% to 0.2% REOs.REE enrichment at Mount Weld is thought to have been caused by


long-term leaching and redeposition by groundwater; it mainly con-sists of REE phosphates (Mariano 1989a), including the yttrium-rich mineral churchite (Lottermoser 1988). During the formation ofthe weathered profile since Permian glaciation, LREEs were incor-porated in phosphate and alumino-phosphate minerals in the upperpart of the profile, and HREEs and Y were deposited as churchite atlower levels (Lottermoser 1990), leading to the formation of bothLREE- and HREE-enriched ore (Figure 8). The Yangibana REE

Adapted from W. Buckovic, personal communication.Figure 5. Stacked cross sections through the Mountain Pass REE carbonatite ore body (gray) and a major associated alkaline rock mass (patterned)

Figure 6. Chondrite-normalized plot of REE distribution (normalized to 100%) in ion absorption clay ores from China and loparite from Russia (data source: Hedrick 1992)

400 NW

800 NW

1200 NW

1600 NW

2000 NW

3000 NW

4000 NW

5000 NW

N

1,000 ft

200 m

Alkaline Rock Outcrop

Carbonatite

Outcrop

Celebration Fault

1,000,000

10,000,000

100,000

10,000

1,000

100

10


Sam

ple:

Cho

ndrit

e

Clay from LongnanClay from XunwuLoparite from LovozeroBayan Obo REE-Fe-fluorite ore(from Figure 4, for comparison)

prospect in northern Western Australia, a gossan containing asmuch as 11% REOs (O’Driscoll 1988), is thought to have beenformed by weathering from a carbonatite associated with Early Pro-terozoic alkaline rocks.

The Araxá carbonatite complex in Brazil contains 800,000 t ofsupergene-enriched laterite that averages 13.5% REOs, mainly inthe phosphate minerals gorceixite and goyazite (Mariano 1989a). Inthe Catalão I Complex about 300 km southeast of Araxá, phosphaticlaterite comprises a resource of about 2 Mt averaging 12% REOs.The REEs are mainly in secondary phosphate minerals, such as gor-ceixite (Mariano 1989a; Morteani and Preinfalk 1996). Laterizedcarbonatite in the Maicuru Complex, Pará, northern Brazil, containsphosphatic material with as much as 17% REEs (Costa et al. 1991).Cerro Impacto, a large circular structure thought to be a deeplyweathered carbonatite that is located in the Amazon region of Vene-zuela, contains as much as 11% REEs in thick laterite (O’Driscoll1988).

Deeply weathered Brazilian carbonatites that contain largetitanium resources in the form of anatase with associated REEphosphate minerals are also a potential source of by-product REEs

Adapted from Duncan and Willett 1990.Figure 7. East–west geological cross section through the Mount Weld carbonatite and its weathered cap

Figure 8. Chondrite-normalized plot of REEs in laterite and carbonatite from Mount Weld, Australia (data source: Lottermoser 1990)

Quaternaryalluvium

Lacustrinesediments

Carbonatiteregolith, residual

Carbonatiteregolith, supergene

Phosphaticregolith

cr

cr

Doleritedike

Glimmerite

Archean basalt

Unweatheredcarbonatite

cu

PREC

AM

BRIA

N

TERT

IARY

(?)

cu

cr

cu

cr

cu

cr

1 km

50 m Depth ofWeathering

1,000,000

100,000

10,000

1,000

100

10


Sam

ple:

Cho

ndrit

e

Fresh carbonatiteLREE laterite oreHREE laterite oreBayan Obo REE-Fe fluorite ore(from Figure 4, for comparison)

NOTE: REE data not normalized to 100% prior to chondrite normalization.


with high LREE/HREE ratios (Mariano 1989a). Titanium concen-trates from a pilot plant contain more than 3% REEs that must beremoved to produce commercial titania. Reserves are known to besubstantial; the largest deposit, at Serra Negra, Minas Gerais, con-tains 200 Mt averaging 27.7% TiO2 (Anon. 1984). Other poorlyexplored carbonatites in the Amazon region of South America areknown to host supergene REE mineralization (Mariano 1989b).

REE-enriched laterites occur at 17 sites in Africa (Orris andGrauch 2002). The most thoroughly investigated deposit is proba-bly monazite-rich laterite that includes 6 Mt of material with about5% REOs that overlies carbonatite at Mrima Hill, Kenya. REEenrichment in the form of secondary hydroxylbastnasite and otherminerals has been reported in karst-related lateritic deposits inEurope, such as upper Jurassic bauxite deposits in Montenegro(Maksimovic and Pantó 1996).

Placer DepositsMost placer accumulations with significant amounts of REE miner-als are Tertiary or Quaternary deposits derived from source areasthat include granitic rocks or high-grade metamorphic rocks; how-ever, paleoplacer deposits that are as old as Precambrian containREE resources. Orris and Grauch (2002) list more than 360 REE-bearing placers. Most commercial deposits are in sands of marineorigin along or near present coastlines and consist of titanium min-eral placers with by-product zircon and monazite; some containxenotime. In the 1980s, monazite and xenotime from titania-zirconpaleobeach placers in Australia were the third most importantsource of REEs in the world, but Australia currently exports little orno REE minerals from such sources, owing to their high thoriumcontent.

Placer rutile-zircon-ilmenite deposits at Eneabba on Austra-lia’s west coast north of Perth, which produced about 2,500 t ofmonazite annually, were brought into production in the late 1970s.These deposits are still productive, but little or no monazite is cur-rently marketed from them. Heavy minerals comprise about 6% ofthe sand mined at Eneabba, and monazite makes up 0.5% to 7.0%of the heavy minerals (Shepherd 1990). The heavy mineral sandslie 30 m or more above present sea level and were deposited in lateTertiary or early Pleistocene time (Lissiman and Oxenford 1975).Archean metamorphic rocks in the interior provided heavy mineralsto Mesozoic sedimentary rocks that were reworked to form thedeposits. In the Geographe Bay area near Capel, ilmenite-zirconpaleobeach deposits that have been mined since the mid-1950soccur along a strike length of about 50 km in Pleistocene to Recentsands (Welch, Sofoulis, and Fitzgerald 1975). The setting andsource are similar to those described for Eneabba. Although theaverage heavy mineral grade of Geographe Bay deposits is some-what higher than at Eneabba, the monazite content of heavy min-eral concentrates is lower, at 0.04% to 1.1% (Welch, Sofoulis, andFitzgerald 1975).

Minor amounts of monazite were once produced from placersalong Australia’s east coast. Rutile and zircon are the main mineralproducts from these operations, which exploit Quaternary beachplacers and sand dunes. Heavy mineral content in some of thedeposits is less than 0.5% (Griffiths 1984), and monazite comprisesless than 1% of the heavy mineral concentrate (McKellar 1975).The WIM 150 deposit, a placer in an interior basin in Victoria, con-tains large REE mineral resources. A 14-m-thick titanium and zir-conium mineral sand deposit, it underlies an extensive areaamenable to dredging. Estimated REE mineral reserves exceed580,000 t monazite and 170,000 t xenotime (O’Driscoll 1988).

In addition to Australia, by-product monazite has beenextracted from beach deposits in Brazil, India, Malaysia, Thailand,

China, Taiwan, New Zealand, Sri Lanka, Indonesia, Zaire, Korea,and the United States. Present-day production is from India, Malay-sia, Sri Lanka, Thailand, and Brazil. In India, monazite productionfrom titania-zircon placers is government controlled, and a domes-tic plant processes monazite concentrate into REE products. Insoutheast Asia, monazite and xenotime are won from concentratesproduced during placer tin, zircon, and titania mining. Prior to1988, when China began to dominate yttrium production, xenotimefrom Malaysia was the largest source of yttrium in the world.Approximately 500 t of monazite per year were produced from1952 to 1994 as a by-product of titania-zircon production fromPleistocene sands near Green Cove Springs in Florida.

Prior to large-scale exploitation of the Mountain Pass deposit,REE production in the United States came almost exclusively fromnonmarine placers, although none comes from such sources now.The Carolina monazite belt, from which a total of about 5,000 t ofmonazite was produced between 1885 and 1917, has considerableplacer reserves that average 0.25 kg/m3 of monazite (Overstreet1967). In Idaho, REE minerals were produced in the 1950s fromplacers. Dredging operations near Cascade produced about 7,000 tof monazite from gravels containing about 1 kg/m3, and monazitereserves are estimated at 38,000 t (Overstreet 1967). Bear Valley,Idaho, where monazite and yttrium-bearing euxenite were minedby dredging, contains an estimated 10,000 t of REOs along withsignificant niobium and tantalum, on the basis of data from Klineet al. (1953). Some Idaho placers contain xenotime, particularlythose in northern Idaho, and REEs are a possible by-product ofplacer gold mining in that area.

Metamorphosed Early Proterozoic conglomerate mined foruranium at Blind River, Ontario, Canada, contains REEs in mona-zite, uraninite, and brannerite (Roscoe 1969), and was the source ofminor REE production from the 1950s to 1970s and again from1986 to 1990. The heavy minerals in this ore are believed to be pri-marily of placer origin with possible enrichment during metamor-phism or hydrothermal activity. Of the approximately 230 Mt of oreaveraging 0.1% U3O8 estimated to be present in the district in 1973(Robertson 1981), perhaps half remained unmined in the early1990s, even though uranium mining didn’t end until 1996.

At Music Valley in Southern California, xenotime- and mona-zite-rich zones with as much as 16% REOs (including 6% Y2O3) inPrecambrian gneiss are thought to have originated as placer accu-mulations (Evans 1964). At Bald Mountain, Wyoming, a paleo-placer monazite resource was identified in Cambrian conglomerate(Borrowman and Rosenbaum 1962).

REE distribution plots of monazite and xenotime from mostplacer deposits have pronounced negative europium anomalies(Figure 9), indicating derivation from sources containing plagio-clase. In addition, placer monazite has relatively high HREE con-tents compared with the strongly LREE-dominated hard-rockdeposits at Bayan Obo and Mountain Pass.

HREE Deposits in Peralkaline Igneous RocksAlthough many REE deposits occur in peralkaline igneous rocksand resources can be large, most such deposits are relatively lowgrade, and only a single deposit in Russia has been mined. Peralka-line REE deposits are typically enriched in yttrium, HREEs, andzirconium. Future development of such deposits is dependent onthe markets for these elements.

REE-bearing loparite has been produced for about 50 yearsfrom nepheline syenites in the peralkaline Paleozoic Lovozeromassif on the Kola Peninsula in Russia. The massif is layered andis composed of four rock suites (Vlasov, Kuz’menko, andYes’kova 1966). The upper 30% of the massif is eudialytic syenite.


The bottom, loparite-bearing section of the massif, over 1,000 mthick, comprises 65% of its volume and consists of alternating lay-ers of nephelinites, nepheline syenites, and associated rocks. Latesyenites occur as relatively small intrusions, and late mafic dikesare rare. The main rock-forming minerals are nepheline, micro-cline, and aegirine with accessory arfvedsonite, hydrosodalite,natrolite, and sodalite.

The loparite mostly occurs as 0.2- to 0.6-mm grains andrarely as larger crystals (Smirnov 1977; Vlasov, Kuz’menko, andYes’kova 1966). Loparite contents are highest in areas of greatestdifferentiation and in rocks enriched in nepheline. It is found pri-marily in porphyritic rocks with concentrations in the urtites andtheir feldspar-aegerine equivalents, juvite, and malignite. The ore,in banded horizons from several centimeters to several metersthick, grades 2% to 3% loparite (Hedrick, Sinha, and Kosynkin1997). The loparite contains 38.5% titanium oxide, 30% to 36%REOs, and 10% to 12% niobium and tantalum oxides. In the1990s, 30,000 tpy of loparite were concentrated locally from orefrom two mines and shipped to Estonia and Kazakhstan for pro-cessing (Kogarko et al. 1995). Loparite mining has ceased inrecent years, but renewed production is being considered. Theloparite REE distribution is weakly enriched in LREEs relative toHREEs (Figure 6).

During the 1980s when yttrium demand exceeded supply,exploration for yttrium-rich REE deposits led to the discovery ofzirconium and HREE-dominated resources that are associated withperalkaline syenitic and granitic rocks. Some have associated beryl-lium, niobium, and tantalum. These resources are large but havelow REE contents compared with commercial LREE-dominateddeposits. Richardson and Birkett (1996) give an overview of Cana-dian peralkaline rock-associated deposits and compare them withsimilar deposits worldwide.

An HREE-zirconium deposit at Pajarito Mountain in NewMexico containing 2.4 Mt of rock with 0.18% Y2O3 and 1.2% ZrO2(Sherer 1990) consists of a 10-km2 dome-shaped syenite intrusion(Kelley 1968). The main zirconium-REE phase was found to beeudialyte (Mariano 1989a).

Foliated syenite at Kipawa Lake, Ontario, includes a 1,300-by-100-m zone of yttrium- and zirconium-rich rock that containseudialyte, mosandrite, and britholite (J. Allan, personal communi-cation). The Y2O3 content of this deposit is comparable to that atPajarito Mountain, but grade varies widely within the deposit,which was intensely deformed during the Grenville orogeny.

The Strange Lake zirconium-HREE-niobium-beryllium depositin Quebec–Labrador is in a circular peralkaline granite complexabout 6 km in diameter (Currie 1985). Reserves are reported at 5 Mtcontaining 0.5% yttrium and 3% zirconium along with beryllium,niobium, and tantalum credits (Miller 1988), and total reserves in thecomplex could be much larger. Most of the REEs are in gadolinite,bastnasite, and kainosite (A.N. Mariano, personal communication).Alkaline complexes in the Shallow Lake and Letitia Lake areas inLabrador, about 250 km southeast of Strange Lake, also includerocks with high yttrium content (Currie 1976; Miller 1988). Granite-and syenite-associated HREE resources with Y2O3 contents of about0.2% have also been identified at Bokan Mountain in southeasternAlaska (Barker and Mardock 1988) and at Thor Lake in the North-west Territories, Canada (Trueman et al. 1988; Taylor and Pollard1996).

A Late Proterozoic peralkaline syenite complex at Ilimaussaqin southern Greenland contains layers with as much as 27% eudia-lyte (Ferguson 1970); the discovery of ore bodies containing 0.12%Y2O3 and 1.2% ZrO2 has been reported (O’Driscoll 1988). Therichest layer is 3.5 m thick and contains about 1 Mt of rock that

averages 4% ZrO2 (Steenfelt 1991). At Norra Kärr, Sweden, Mid-dle Proterozoic eudialyte-bearing peralkaline syenite has been eval-uated as a source of zirconium (Lundqvist 1980). Eudialyte fromNorra Kärr contains 1.3% yttrium (Fryer and Edgar 1977).

Four Saudi Arabian deposits carrying REEs and zirconiumoccur in alkaline granites. The deposits include both LREE- andHREE-rich types, and estimated sizes range from 6 to 440 Mt withyttrium contents from 0.13% to 0.52% (Drysdall et al. 1984). TheBrockman deposit in northern Australia has been evaluated for zir-conium, yttrium, HREEs, niobium, and tantalum. The deposit, inmetamorphosed lower Proterozoic rhyolite tuff, consists of 9 Mtcontaining 1.3% zirconium, 0.15% Y2O3, and 0.12% HREEs(O’Driscoll 1988). A deposit of altered Jurassic intrusive trachyte atDubbo, New South Wales, Australia, is probably in this class. Itcontains 83 Mt with 1.9% zirconia, 0.14% Y2O3, 0.12% REOs, andtantalum credits (Australian Zirconia Ltd. 2000). At Pocos de Cal-das, Brazil, a Mesozoic or Early Tertiary peralkaline syenite com-plex 30 km in diameter (Woolley 1987) contains localconcentrations of eudialyte. On the Kola Peninsula, Russia, Paleo-zoic peralkaline syenite complexes at Khibina and Lovozero con-tain large amounts of eudialyte-bearing rock (Gerasimovsky et al.1974; Kogarko et al. 1995).

Vein DepositsREE deposits in veins are typically small in comparison to the com-mercial hard-rock deposits at Bayan Obo and Mountain Pass. Nev-ertheless, REEs were produced from two vein deposits in Africa inthe past, and more recently from two such deposits in China.

In recent years, Chinese vein deposits have become importantREE sources. In 1992, production from the Maoniuping mine,Sichuan Province, and the Weishan mine, Shangdong Province,together accounted for 24% of Chinese REE production (Wu, Yuan,and Bai 1996). Both are described as bastnasite-barite-carbonateveins associated with quartz syenite. They may be carbonatitedikes; published descriptions are not conclusive. The Maoniupingdeposit, which is reportedly the second largest REE deposit inChina (Pu, Chen, and Yang 2001), consists of swarms of veins andadjacent veinlets as much as 1,000 m long and 20 m wide that aver-age about 2% REOs (Wu, Yuan, and Bai 1996). The ore mineral iscoarse-grained bastnasite. Other REE minerals include chev-kinite, xenotime, britholite, allanite, and monazite. Gangue miner-als are barite, calcite, quartz, fluorite, feldspars, aegerine-augite,

Figure 9. Chondrite-normalized plot of REE distribution (normalized to 100%) in placer monazite and xenotime (data source: Hedrick 1992)

1,000,000

100,000

10,000

1,000

100

10


Sam

ple:

Cho

ndrit

e

Monazite from Capel, AustraliaMonazite from Guangdong, ChinaXenotime from MalaysiaBayan Obo REE-Fe fluorite ore(from Figure 4, for comparison)


and sulfide minerals. Ore at the Weishan mine occurs in veins asmuch as several hundred meters long and 1 m wide, with an aver-age grade of 1.6% REOs. The mineral assemblage is similar to thatat Maoniuping but also includes dolomite, amphiboles, thorite, tita-nium minerals, and niobium minerals (Wu, Yuan, and Bai 1996).

In South Africa, more than 50,000 t of monazite were pro-duced in the 1950s and 1960s at Steenkampskraal (Neary and High-ley 1984) from a monazite-apatite-quartz vein in Proterozoicgranite. The vein, which occupies a shear zone, is about 300 m longand 1 m thick (Andreoli et al. 1994). Monazite comprises as muchas 75% of the vein material (Overstreet 1967), accompanied by apa-tite, magnetite, and sulfide. The ore contains as much as 39% REOswith LREEs dominating HREEs and a pronounced negativeeuropium anomaly (Andreoli et al. 1994). A similar monazite-apatite vein cuts Precambrian granite near Crescent Peak in Nevadaabout 30 km east of the Mountain Pass REE deposit (Castor 1991).

At Karonge, Burundi, in east Africa, quartz-barite-bastnasite-monazite stockwork veins that cut Precambrian quartzite and schisthave produced small amounts of REEs (Notholt, Highley, andDeans 1990). The bastnasite and monazite have high LREEs typicalof carbonatites, but secondary cerianite and rhabdophane haveabnormally high or low cerium contents because of oxidation (VanWambeke 1977). In South Africa, small REE reserves have beenestimated for britholite-bearing veins in the Pilanesberg peralkalinecomplex (Lurie 1986).

REE- and thorium-bearing quartz veins scattered over an areaof about 250 km2 at Lemhi Pass on the Idaho–Montana border inthe United States comprise a REE resource of 370,000 t (Staatz,Sharp, and Hetland 1979). The mineralization is mostly LREE-dominated and averages less than 1% REOs, but some veins con-tain as much as 0.3% Y2O3 and 2.0% total REOs. A small area atDiamond Creek about 50 km northwest of Lemhi Pass containsveins with similar mineralogy but with low LREE/HREE ratios.The Snowbird occurrence in northwestern Montana is an REEresource of unknown size with 0.1% to 0.2% Y2O3. The REE min-erals are in a Cretaceous quartz-carbonate-fluorite-parisite veinconsidered to be of hydrothermal origin (Metz et al. 1985). Thorite-bearing quartz-carbonate-fluorite veins at Hall Mountain near Port-hill in northernmost Idaho contain as much as 0.2% Y2O3. The Ter-tiary Bear Lodge trachyte-phonolite complex in Wyoming containsa minor REE resource in veins that average about 3% REOs (Staatz1983), but the veins are narrow and commercial development isunlikely.

Veinlike deposits of apatite at Nolan’s Bore, near AliceSprings in Australia, contain a significant amount of REEs. Aninferred resource of 3.8 Mt of ore contains 4.0% REOs and 17%P2O5 (Anon. 2003). The REEs reportedly occur in the apatite and inREE minerals, with the latter possibly occurring as cross-cuttingzones of cheralite.

Other DepositsSome REE accumulations that do not fit into the deposit types pre-viously discussed have had REE production or have been evaluatedas possible sources. In recent years, a deposit near Aktyus in theTien Shan Range of Kyrgyzstan has seen significant REE produc-tion. Although little information on this deposit exists in the litera-ture, the deposit is said to contain complex ore from which lead,molybdenum, silver, and bismuth are also produced (GeologicalSurvey of Kyrgyzstan 2004). The average REE content of the ore isreported to be 0.25%, with yttrium and HREEs making up 43.7% ofthe total REEs. The deposit consists of two stock-shaped bodiesthat contain synchisite-(Y), bastnasite, monazite, xenotime, andzircon (Roskill Information Services Ltd. 1996).

Some fluorspar deposits offer potential for REE productionfrom associated REE minerals or from REEs that substitute for cal-cium in fluorite. Fluorspar mining at Naboomspruit in South Africaproduced a small amount of monazite in the 1980s (Hedrick andTempleton 1991). About 65 t of bastnasite concentrate were pro-duced from a fluorspar deposit in the Gallinas Mountains, NewMexico, in the 1950s (Adams 1965).

Near-economic HREE-uranium mineralization occurs asxenotime accumulations in sandstone of the Late PrecambrianAthabasca Group near Wheeler River in Alberta, Canada. Similaraccumulations of HREEs and uranium are in quartzite in WesternAustralia (Mariano 1989a).

REE deposits in Vietnam are reported to contain several mil-lion tons with grades that range from 1.4% to 5% REOs withLREE- to moderate HREE-enriched distributions. The depositshave been described as crushed zones in Paleozoic limestoneaffected by metasomatic processes (Premoli 1989).

Marine phosphorites have been proposed as a potential sourceof REEs (Altschuler, Berman, and Cuttitta 1967). Certain membersof the Permian Phosphoria Formation, which are mined for phos-phate in large quantities in Montana and Idaho, contain significantamounts of REEs, including as much as 0.1% yttrium (Gulbrandsen1966).

Pegmatite REE deposits are common but generally too smallor too low in grade to be commercially exploited. Production ofREEs from pegmatite mined for other minerals, such as feldspar ormica, however, is possible. Allanite-bearing pegmatites are rela-tively common, generally with high LREE/HREE ratios. In Austra-lia’s Northern Territory, a pegmatite deposit containing 1 Mt with4% allanite was investigated as a source of REEs (O’Driscoll1988).

ORIGIN OF DEPOSITSThe concentration and distribution of REEs in natural deposits aredependent on several petrogenetic processes, including enrichmentand complexing in late-stage magmatic or hydrothermal fluids,fractionation into mineral phases, oxidation or reduction, and redis-tribution during weathering. LREE enrichment in igneous rocksgenerally is ascribed to fractionation of HREE into minerals such asgarnet and pyroxene during partial melting of source materials orduring fractional crystallization. In addition, anomalously loweuropium/chondrite contents, a common feature in highly evolvedigneous rocks of crustal derivation, are considered to be the resultof Eu fractionation into plagioclase. Eu/Eu* (measured europiumdivided by europium calculated from interpolation between samar-ium and gadolinium; Henderson 1984) generally is near unity inmantle-derived rocks such as carbonatites but low in granitic rocksthought to have been derived by partial melting of crustal materials.

Lanthanum/gadolinium (La/Gd), as a measure of LREEs/HREEs, plotted against Eu/Eu* (Figure 10) is helpful in the classi-fication of REE deposits and the interpretation of their origin.Whole rock and mineral plots for carbonatite REE deposits and lat-erite derived from carbonatite fall in a field of high La/Gd and Eu/Eu*. Carbonatites are considered to be mantle-derived rocks withlittle or no contribution from the crust. By comparison, monaziteconcentrates from Australian, Chinese, and Florida placer depositsplot in a tight cluster with high La/Gd and low Eu/Eu*, probablyreflecting derivation from rocks that contained plagioclase or wereproduced by partial melting of plagioclase-bearing crustal rocks.HREE deposits in peralkaline rocks plot in a diffuse field with rela-tively low La/Gd and Eu/Eu*, probably because of derivation of thehost peralkaline rocks from plagioclase-bearing crustal sources,and placer xenotime from Malaysia plots near the peralkaline field.


Some iron-REE deposits occupy a linear field that appears to con-nect the carbonatite and peralkaline HREE deposit fields, suggest-ing mixed crustal and mantle sources for the REEs, whereas orefrom the Bayan Obo deposit plots within the carbonatite field, sug-gesting mantle derivation.

Most significant natural concentrations of REEs in nature arein, or associated with, alkaline igneous rocks and carbonatites. Fordeposits such as those at Mountain Pass, Mount Weld, Araxá, PeaRidge, and Lovozero, this association is clear. At Bayan Obo,which emerged as the world’s most important source of REEs in the1990s, the association is less clear. Some researchers believe thatthe Bayan Obo deposits originated as carbonatite (Bai and Yuan1985; Le Bas, Spiro, and Yang 1997). Others believe that thedeposit is hydrothermal, but call on an alkaline igneous or carbon-atite source for the fluids (e.g., Drew, Meng, and Sun 1990; Yuanet al. 1992; Campbell and Henderson 1997). Still others call onmobilization of lower crustal REEs and deposition by hydrothermalactivity during a protracted period of subduction (Chao et al. 1997).In addition, the Bayan Obo iron deposits have been compared tobanded-iron formation deposits (Qiu, Wang, and Zhao 1983).

Hitzman, Oreskes, and Einaudi (1992) proposed a class ofdominantly Proterozoic iron oxide (copper-uranium-gold-REE[Cu-U-Au-REE]) deposits that includes the Bayan Obo, OlympicDam, Kiruna, and Pea Ridge deposits. They suggested that thesedeposits formed at relatively shallow crustal levels (< 4 to 6 km)from igneous-hydrothermal systems tapped by deep crustal struc-tures associated with global rifting, possibly during break-up of aProterozoic supercontinent. Hitzmanm, Oreskes, and Einaudi(1992) did not call for mantle or carbonatite-related fluids for thisclass of deposits, instead proposing release of large volumes ofmagmatic-hydrothermal fluid during magmatic underplating of thecrust. By contrast, Chao et al. (1997) proposed a protracted Paleo-zoic lower crustal origin for REE-rich fluids at Bayan Obo on thebasis of ages of amphiboles related to REE deposition. Smith(2001) noted evidence that the Olympic Dam deposit was formedfrom hypersaline brines and proposed that it should be placed in aseparate category from Bayan Obo, which shows no evidence ofsuch fluids.

The Mountain Pass deposit shares features with other carbon-atites in the world: (1) textural and structural features that supportigneous intrusive origin (Olson et al. 1954); (2) associated alkali-rich fenitic alteration; and (3) low 87Sr/86Sr (Powell, Hurley, andFairbairn 1966; E. DeWitt, personal communication). But the shapeof the carbonatite ore body, its ultrapotassic alkaline association(most carbonatites are associated with alkaline rocks that are domi-nantly sodic), and chemistry (it is unusually low in iron, phospho-rus, and niobium for REE-rich carbonatite) suggest that its sourcewas different than that of most other carbonatites (Castor andNason 2004). The chemistry of the ultrapotassic alkaline rocks atMountain Pass suggests that they were derived from primitive ordepleted mantle mixed with an enriched mantle or crustal compo-nent, and the carbonatite likely came from a similar source. TheMountain Pass LREE deposit occurs in a northeasterly-trendingbelt of anorogenic Middle Proterozoic plutonism that crosses theNorth American continent (Anderson 1983) and contains eightMiddle to Late Proterozoic REE deposits (Castor 1993, 1994).

Although there are mineral and chemical similarities betweenthe Mountain Pass and Bayan Obo deposits, there are also majordifferences. Like the Bayan Obo ore, the Mountain Pass carbonatiteore has extreme LREE enrichment with no europium anomalies(Figure 4) and low 87Sr/86Sr. As at Bayan Obo, alkali-rich alterationis associated with the Mountain Pass carbonatite; in both areas,alteration assemblages include sodic amphibole and potash feldspar

(Olson et al. 1954; Drew, Meng, and Sun 1990). The Bayan Oboand Mountain Pass deposits are both enriched in barium and fluo-rine (Wu, Yuan, and Bai 1996; Castor and Nason 2004). Samples ofBayan Obo ore, however, have an average BaO content of about2.4%, and a maximum barium content of 7.7% (based on data fromChao et al. 1997), much lower than Mountain Pass ore (Table 5).The fluorine content of Bayan Obo ore, which averages more than9%, is much higher than that of Mountain Pass ore, and Bayan Oboore also has higher phosphorus and niobium contents (Table 5).Deposits such as those at Bayan Obo, Pea Ridge, and Mineville areclearly different from Mountain Pass in their association with largeamounts of iron. Whereas Mountain Pass ore contains about 10times as much bastnasite as monazite, Bayan Obo deposits containapproximately equal amounts of bastnasite and monazite (Y. Qiu,personal communication), although bastnasite-rich ore is selec-tively mined. REE-chondrite patterns for Bayan Obo show enrich-ment in HREEs relative to Mountain Pass (Figure 4). Finally, theultrapotassic alkaline intrusions that are associated with the Moun-tain Pass carbonatite are not found at Bayan Obo. On the basis of itsrelatively high P, Fe, and Nb contents, if the Bayan Obo deposit hasa carbonatite-related origin, it is probably to a nephelinite-carbon-atite system rather than to an ultrapotassic-carbonatite system as atMountain Pass.

The chemical compositions and mineral assemblages of mostvein-type REE deposits suggest a genetic relationship with carbon-atite-alkaline complexes. Such a relationship has been proposed for

Figure 10. Plot of La/Gd against Eu/Eu* for some REE-rich mineral concentrate and rock samples (data sources: Deans 1966; Fryer and Edgar 1977; Qiu, Wang, and Zhao 1983; Drysdall et al. 1984; Hedrick 1985; Roeder et al. 1987; Barker and Mardock 1988; Nelson et al. 1988; Mariano 1989a; Costa et al. 1991; G. Morteani, personal communication; L. Tucker, personal communication)

La/G

d

1,000

100

10

0.1

0.01

Eu/Eu*

1

0.01 0.1 1

PlacerMonazite

Iron-REEDeposits

Carbonatites

PeralkalineREE Deposits

21

10

11

12

13

14

16 17

1819

20 15

1A

1B

2 3

4A5

67

89

4B

Deposit Name and Type

South China placer monaziteFlorida placer monaziteBrazil placer monaziteMalaysia placer xenotimeStrange Lake peralkaline igneous HREE depositsKipawa Lake eudialyteBokan Mountain peralkaline igneous HREE depositsJabal Sa'id and Jabal Tawlah peralkaline igneous HREE depositsCrescent Peak vein deposit

1213141516

1718

19, 20

21

LocationNo.Deposit Name and Type

Mountain Pass bastnasiteMountain Pass monaziteKangankunde REE-rich carbonatiteKizilçäoren REE-rich carbonatiteSalitre II perovaskite and anataseMount Weld carbonatiteMaicuru REE-rich carbonatiteBayan Obo orePea Ridge brecciaMineville apatiteWestern and Eastern Australia placer monazite

1A1B23

4A, 4B56789

10, 11

LocationNo.


the Karonge, Burundi, veins (Van Wambeke 1977). The monazite-apatite vein at Crescent Peak, Nevada, though, has REE distribu-tion that includes a prominent negative europium anomaly that sug-gests a different origin than the Mountain Pass deposit 30 km away.

Although REE minerals are considered to be relatively unaf-fected by weathering, some lateritic REE deposits show signs ofmobilization and redeposition of REEs. Residual concentration ofREE minerals such as monazite is possible in laterite, and Morteaniand Preinfalk (1996) reported that the REEs behaved as rather immo-bile elements in laterites at Araxá and Catalão in Brazil. Differencesin REE distributions in different lateritic deposits may result fromdifferences in parent rocks. For instance, Morteani and Preinfalk(1996) noted that REE distributions in the Araxá and Catalão lateritesmirrored those in underlying alkaline rocks, and the very differentREE distributions in the Xunwu and Longnan deposits in southernChina (Figure 6) may be caused by different REE distributions in theparent rocks. As noted by Mariano (1989a), however, development ofREE mineralization in laterites does not require derivation from inde-pendent REE minerals in the parent rock but may come from REEsin minerals that are less resistant to breaking down during weather-ing. Fractionation of the REE elements clearly takes place duringweathering, as indicated by differences in REE distributions at differ-ent levels in the weathering cap above the Mount Weld carbonatite(Lottermoser 1990; Figure 8). The production of cerium anomalies(either positive or negative) during the breakdown of REE-bearingminerals in an oxygen-rich environment is to be expected; however,laterite deposits may or may not show such fractionation (compareFigure 8 with Figure 6). Mariano (1989b) proposed that differencesin the mineral assemblages in carbonatite-derived laterites may resultfrom factors other than parent rock composition, such as age, depthof weathering, and geomorphology.

TECHNOLOGYExploration TechniquesGeologic conditions favorable for REE deposits are widespread,but most of the world’s hard-rock deposits are restricted to areasunderlain by Precambrian rock. The Bayan Obo deposits, which arein Proterozoic rock, were originally discovered as ridge-formingsurficial deposits of dark brown and black iron oxide (Argall 1980).Carbonatite REE deposits and many HREE deposits occur in or areassociated with alkaline rocks in circular complexes, generally inPrecambrian host rocks. Such complexes may be identified usingaerial photography, even in deeply weathered and highly vegetatedareas. Although isolated alkaline-carbonatite complexes such asMount Weld in Australia do occur, most are in clusters or linearbelts like those associated with rift zones in eastern Africa and theAraxá–Catalão belt in Brazil.

Sources for placer REE deposits include granitic rocks orhigh-grade metamorphic rocks, or both. Coastal monazite-bearingplacers in Western Australia, Brazil, India, Sri Lanka, and thesoutheastern United States were derived from highly metamor-phosed Precambrian shield rocks; whereas monazite and xenotime-producing placers in Eastern Australia, Malaysia, Indonesia, China,and Korea were eroded from Phanerozoic granites, including tin-rich granites. Sources of monazite, euxenite, and xenotime in Idahoplacers include both highly metamorphosed Precambrian rocks andMesozoic granitic rocks.

Because REEs are associated with thorium and uranium, radi-ometric exploration techniques are extremely useful in REE explo-ration. REE-rich carbonatite at Mountain Pass was discoveredduring surface prospecting for uranium using a geiger counter(Olson et al. 1954). Many other REE deposits were found by sur-face or airborne radiometric surveying. Although most placer REE

deposits yield only subtle radiometric signatures, careful data col-lection and analysis can be used to locate favorable intrabasin areasor horizons.

Geochemical prospecting using heavy-mineral concentrationof active stream sediments can be very effective because most REEminerals are relatively heavy and resistant. Concentrates made byhand panning, mechanized gravity concentration, or heavy mediaseparation can be analyzed for REEs using inexpensive multiele-ment analytical techniques. Because many REE minerals are resis-tant to chemical breakdown, use of analytical methods that do notrequire dissolution, such as x-ray fluorescence or neutron activationanalysis, is desirable, or care must be taken to use methods thatassure total dissolution when using a technique such as inductivelycoupled plasma (ICP) spectroscopy. During prospecting for REEplacer deposits in Idaho, the senior author found that favorablebasins yield heavy-mineral concentrates with 1% or more ceriumand 0.1% or more yttrium. Splits of heavy-mineral concentratestaken during exploration should be saved so mineral determinationscan be performed using microscopic or other techniques.

Regional geochemical surveys can be used to find igneousrocks that are likely to host or be associated with REE deposits.Carbonatite deposits were delineated in Greenland using multiele-ment analysis of fine stream sediment samples collected at a den-sity of 1 sample/30 km2 (Steenfelt 1991). The Magnet Covealkaline-carbonatite complex in Arkansas was delineated usingREEs, titanium, and fluorine contents of stream silt samples col-lected at a density of 1 sample/10 km2 (Sadeghi and Steele 1989).Peralkaline rocks that host HREE deposits in Greenland weredelineated using niobium content in fine stream sediments col-lected at 1 sample/6.25 km2 (Steenfelt 1991). The Strange Lakedeposit in Canada was discovered using regional lake water andsediment surveys, with subsequent tracing of glacially transportedboulders that were initially recognized 20 km from their source(Richardson and Birkett 1996).

Biogeochemical prospecting may be useful in defining buriedREE targets. REE contents of plant ash have been analyzed in areasthat contain pegmatitic and vein occurrences in Finland and Can-ada, and extreme LREE enrichment has been noted in plant ashfrom the Bayan Obo area (Dunn 1995).

Normal surface exploration techniques, such as trenching andpitting by heavy equipment, are used to evaluate poorly exposedREE deposits. Exploratory dredging is used to locate beach placerdeposits. In countries with low labor costs, hand pitting may beused to evaluate placer deposits or unconsolidated deposits such asthose in laterite.

Drilling techniques utilized for hard-rock REE exploration andprospect evaluation are mainly core drilling and dual-tube rotary orhammer drilling. The method of choice for most heavy-mineralplacer evaluations is dual-tube rotary drilling. The collection of largesamples, such as are necessary to evaluate gold placers, generally isnot needed because the nugget effect is not a factor in most heavy-mineral sands that contain REEs.

Because of the economic importance of differences in pro-cessing REE-bearing minerals, identification of the REE-bearingphases in prospective deposits is important. Light-colored REEminerals, such as bastnasite and monazite, may generally be distin-guished by a green luminescence when illuminated by a mercuryvapor light (Murata and Bastron 1956). A conventional black lightwith the purple filter removed may be used for this test. This lumi-nescence, which is caused by absorption and indicates the presenceof significant amounts of neodymium, has been used successfullyfor preliminary REE-grade determination during core logging atMountain Pass. Because the test is based on neodymium content,


HREE minerals may not respond. Absorption bands in the visiblespectrum may also be used to distinguish REE minerals with theaid of a hand spectroscope (Mertie 1960).

Geophysical methods that can be used to search for buriediron-REE deposits and some REE carbonatite deposits includegravity surveys and airborne or surface magnetic surveys. Theeffectiveness of such methods is based on anomalous density andmagnetic susceptibility of a deposit and the associated rocks.Because many carbonatite complexes have a central carbonatitemass surrounded by mafic alkaline rocks, a common signature is amagnetic bull’s-eye with high central values combined with a cen-trally located gravity low and a concentric gravity high. The Moun-tain Pass carbonatite, however, does not follow this pattern becauseof its high density (caused by abundant barite and bastnasite) andlack of magnetite. The Mount Weld carbonatite was discoveredduring interpretation of a regional aeromagnetic survey, and itsdetailed magnetic expression exhibits concentric zoning (Gunn andDentith 1997). Other geophysical techniques include seismic sur-veys, which may be used in conjunction with drilling to determinethe depth to bedrock of placer deposits.

MiningAt Mountain Pass, REE ore was mined until October 2001 in anopen pit approximately 150 m deep using 80-t haul trucks and a10-m3 shovel. Blast holes drilled at 3 to 4 m spacing were assayedfor total REOs and other elements by x-ray fluorescence methods.Approximately 300,000 tpy were mined with a stripping ratio of5:1 or higher. The ore was crushed using jaw and impact crushers atthe upper lip of the pit and fed to an adjacent mill. Mining opera-tions may be restarted in 2006 following governmental approval forpit expansion and a new tailings disposal site; however, newlyenacted California mining regulations could block the resumptionof mining.

At Bayan Obo, iron and REE ore is mined from two largeopen pits at a rate of at least 15,000 tpd using electric shovels andrail haulage (Argall 1980). Because water is lacking near the mine,crushed ore is hauled about 150 km by railroad to milling facilitiesat Baotou. According to Argall (1980), there was no selective min-ing at Bayan Obo to maximize production of any group of minerals.Originally, REO minerals may have been recovered from iron orewaste. According to Jackson and Christiansen (1993), however,REO in the iron ore is locked in slag during refining and is notrecovered, and REO is produced strictly from bastnasite ore thatoccurs in certain zones. According to L. Drew (personal communi-cation), during the late 1980s, REE ore came from areas mined spe-cifically for REE production, although iron ore and REE ore werehauled on the same trains to Baotou, where most cars were routedto steel mills and some to nearby REE processing facilities. Duringthe 1990s, REE ore enrichment by hand cobbing was witnessed atthe mining operation.

Russia’s loparite was mined as a primary product by under-ground and open-pit methods from ore bodies in the Lovozeroalkali massif near the city of Revda. The mining company, Lovo-zersky GOK, has plans to reopen and expand its mining capacity to2.6 Mtpy of ore; however, problems with downstream processingof loparite concentrate in Estonia probably precludes expansion inthe near future.

REE placer mining is done by either dry land mining or bydredging techniques. Although Australia does not currently exportmonazite from titania and zircon placer operations, descriptions ofthe placer mining operations in that country are probably represen-tative of placer mining that yields monazite and xenotime else-where. Dredges with capacities of as much as 2,800 tph (Anon.

1987) were used in the 1980s to mine beach and dune sand depos-its, particularly along the east coast. Dry land techniques, includingloader and truck, scraper, drag line, or bucket wheel mining, areused for most of the heavy mineral sand mining in Western Austra-lia (Griffiths 1984).

ProcessingFlotation was used at Mountain Pass to make a bastnasite concen-trate containing about 60% REOs until late in 2001. Figure 11 is asimplified flow chart for concentrate production at Mountain Pass.This concentrate was used either (1) on site as feed for chemicalseparation of REEs; (2) leached with dilute HCl to produce a 70%REO concentrate; or (3) shipped as is. Current sales are from stock-pile. Flotation that produces concentrate containing about 60%REOs is used at Baotou to process REE ore from the Bayan Obodeposit. In Russia, loparite ore was processed using gravity andelectromagnetic separation methods to produce a 95% loparite con-centrate. The mill has the capacity to produce about 6,500 t of con-tained REOs in loparite concentrate annually, and an expansion to12,500 t of capacity is planned.

REE minerals are separated commercially from associatedminerals in placer deposits using a combination of gravity, mag-netic, and electrostatic techniques (Griffiths 1984). Figure 12 is atypical beneficiation flow diagram for monazite and xenotime.Gravity methods include the use of jigs, spiral and cone concentra-tors, and shaking tables. Sizing and preconcentration commonly isperformed at the mine site by trommels, shaking screens, and grav-ity separation. Many dredges have such facilities on board or utilizefloating preconcentration plants.

REE extraction from monazite and xenotime is accom-plished by dissolution in hot concentrated base or acid solutions.During past processing of monazite ore, REEs were extractedusing a concentrated solution of sodium hydroxide at 140° to150°C (Kaczmarek 1980). After cooling, hydroxides of REEs andthorium were recovered by filtration, and thorium was separatedby dissolution and selective precipitation. Monazite and xenotimealso have been processed using hot sulfuric acid digestion andwater leaching to remove phosphate. This is followed by selectiveprecipitation of thorium during dilution and precipitation of REEsas double sulfates.

At Mountain Pass, bastnasite was calcined to drive off CO2and fluorine and leached with HCl to dissolve most of the trivalentREEs (Figure 11). The residue, which consists mostly of CeO2, wassold as a polishing abrasive. At Baotou, Bayan Obo REE mineralconcentrate is baked with sulfuric acid at 300°C to 600°C andleached with water, taking REEs into solution and precipitatingother elements as waste (Chegwidden and Kingsnorth 2002). REEsare then precipitated as double sulfates and converted to hydrox-ides, which are leached with HCl for purification using solventextraction (SX) and other methods. Because this method is verysimilar to placer monazite processing, the Bayan Obo concentrateis assumed to contain significant amounts of monazite.

Russian loparite concentrate is produced by a mill with thecapacity to generate about 6,500 tpy. The concentration processuses gravity and electromagnetic separation methods, and yields aconcentrate containing 95% loparite. The mill also produces aegir-ine and nepheline-feldspar concentrates. If mining expansion plansmove forward, expansion of the mill will produce 3,000 t ofloparite concentrate in a pilot phase followed by 12,500 t in thecommercial phase. Additional expansion, if warranted, wouldincrease loparite concentrate capacity to 25,000 tpy.

Russian loparite concentrate is processed using gaseous chlo-rination at high temperature in the presence of reducing agents


Source: Hedrick 1992.Figure 11. Generalized bastnasite beneficiation flow diagram (Mountain Pass, California)

Source: Hedrick 1992.Figure 12. Generalized flow diagram for extraction of monazite and xenotime from Ti-Zr-REE mineral sand

Bastnasite Concentrate(60% REO)



TailingsTailings

SX

Ce Concentrate(57% CeO2)

Ca and Sr CO2

Ce Carbonate(90% CeO2)

LnCl3 w/42% CeO2 LnCl3 w/15% CeO2

Miningby Bench-Cut,

Open-Pit Methods

Jaw Crusher(<10 cm) Screen Cone Crusher

(<9.5 mm)Ball Mill

(<150 µm)

Rotary KilnDryer

Hydrochloric AcidLeach

Roaster(Oxidizing)

Thickenerand Filter Redox

Drum Filter ThickenerFlotation Cells

(Cleaner,Scavenger)

Flotation Cells(Rougher)

Conditioning Tank(Chemicals and

Steam)

LnCl3w/CeO2

Tailings

Oversize

Tailings

Middlings

Tailings

Tailings

Con

c.Middlings

Middlings

Tailings

Mags

Non-Mags

Ilmenite(Natural Mags)

Non-MagsConductors

Nonconductors

InducedMags Non-Mags

Conductors

Nonconductors

Nonconductors

NonconductorsNon-Mags

Zircon

Nonsulphate Ilmenite(Induced Mags)

Altered Ilmenite(Induced Mags, Conductors)

(Non-Mags,Nonconductors)

Leucoxene(Non-Mags, Conductors)

Induced MagsNonconductors

Non-MagsConductors

Non-MagsConductors

Non-MagsConductors

Monazite (MediumInduced Mags,Nonconductors)

Xenotime(MediumInduced Mags,Nonconductors)

Mining byDredging or

ScrapingScreens Primary

SpiralsSecondary

SpiralsMid-Cleaner

SpiralsConc.

Primary MagneticSeparators

Induced RollMagnetic

Separators

SecondaryMagnetic

Separators

ElectrostaticPlate Separators

Conc.

FinishingSpirals

Conc.Wash

(Remove Organicsand Coatings)

Rinse andDewaterRotary Dryer

High-TensionRoll Separators


Induced RollSeparators

Induced RollMagnetic Separators




Induced RollSeparators

Wet Table orAir Table


Figure 13. Loparite refinery flow diagram (Irtysh, East Kazakstan Oblast, Kazakhstan, and Sillamjae, Estonia)

Decomposition

Prec

ipita

teD

isso

lutio

n

H2SO4 Ti, Nb, and TaSolution

Cl (gaseous) H2O Na2CO3

Ti, Nb, and TaVolatiles

FusionCake

LopariteConcentrate

Chlorination at750˚–850˚C

Solution

Precipitate

Decomposition with85% H2SO4

at 200˚C with(NH4)2SO4

Rare-Earth andThorium Sulfate

HNO3

Rare-Earth andThorium Carbonate

Rare-Earth andThorium Nitrate

ThoriumConcentrate

Na4OH

Rare-EarthNitrate

Rare-EarthSX Separation

LanthanumConcentrate

LanthanumSX

CeriumConcentrate

Ce Oxide,Ce Fluoride,

Polishing Compounds

Nd, Pr, Sm, EuConcentrate

Nd, Pr, Sm, EuSX

Precipitate

(Mikhaillichenko, Mikhlin, and Pattrikeev 1987). The more volatilechlorides of titanium, niobium, and tantalum are separated from theless volatile chlorides of REEs and other elements, which remain asa fusion cake (Figure 13). The fusion cake is dissolved in hot sulfu-ric acid in the presence of ammonium sulfate (Kosynkin et al.1993). The solution is diluted with water, dropping out double sul-fates of rare earths and thorium that are converted to carbonates bythe addition of sodium carbonate. The carbonate is dissolved innitric acid, and thorium is precipitated by raising solution alkalinity(Hedrick and Sinha 1994) or by solvent extraction. The remainingrare earth nitrate solution is separated and purified by selective pre-cipitation and solvent extraction.

Because of their chemical similarity, the trivalent REEs are dif-ficult to separate. Although fractional crystallization and ion-exchange techniques are used to separate them in small amounts,commercial separation generally is done using liquid–liquid solventextraction (Kaczmarek 1980). This process consists of addition of asolvent composed of a mixture of organic compounds to the pregnantaqueous solution in a series of mixing/settling cells that allow repeti-tive fractionation during a more-or-less continuously flowing pro-cess. Figure 14 is a simplified diagram of the SX process used atMountain Pass. Following precipitation and drying, specific REEcompounds with purities in excess of 99.99% can be produced bythis process. At Mountain Pass, a REE fraction produced during SXwas treated to separate europium by reduction to the divalent state,with ultimate purification as europium oxylate (D. Witham, personalcommunication). Other individual REE compounds were precipi-tated as hydroxides, carbonates, and oxylates following complex iter-ative SX processes.

Specifications and TestingSpecifications for REE mineral concentrates, compounds, and met-als vary depending on use. Purity specifications for some com-pounds may be determined by reference to Table 6. Testing of REEproducts mainly consists of purity determinations. In addition tospecific wet chemical analytical procedures, the most effectivetechniques used to test products for individual REE content arex-ray fluorescence, ICP emission spectroscopy, and instrumental

neutron activation analyses. Impurities, such as phosphate, silica,lead, and thorium, may be deleterious to product utilization or theenvironment. The need to limit such impurities depends on use.Physical specifications, such as particle-size distribution, color, andmoisture content, also are important for some products.

Advances in processing in recent years have removed marketsfor some relatively low-spec REE products. For instance, a 90%cerium oxide product used mainly by the glass industry and manu-factured cheaply for years at Mountain Pass has been supplanted byhigher-grade cerium products with little or no price increase.

ECONOMIC FACTORSPricesREE mineral concentrates and intermediate compounds had rela-tively steady price and production increases from the 1960s throughthe 1980s, when markets were created in response to new technolo-gies (Castor 1994), while REE sources remained relatively steady.In recent years, prices for REE concentrates and intermediate com-pounds can be considered inexpensive when compared with manyother mineral products (Table 6) and have lagged behind inflation.REE production increases accelerated following significant Chi-nese impact on the market in the 1990s (Figure 1), while prices forall REE commodities have decreased (Table 7). Therefore, despite atotal annual production increase on the order of 300% since theearly 1980s, the overall dollar value of the REE market has proba-bly remained static or even decreased.

Prices for individual REOs and metals were generally lower inthe late 1990s and early 2000s than in the 1960s, 1970s, and 1980s.Individual REE compounds vary widely in value (Table 7), mostlybecause of relative abundance and production costs. The mostexpensive REE is the heaviest—lutetium, which sold as the high-purity oxide for as much as $9,000/kg in 1989 (Hedrick 1991a) anddeclined to a low of $2,400/kg in 2003. Metallic REEs typicallyhave higher prices than their equivalent oxides or other compounds.Mischmetal, a mixture of metallic REEs, was quoted at $12.35/kgat the end of 1989 (Hedrick 1991a). During the 1980s, individualREE prices generally were static or increased slightly, but, sincethen, market changes have caused price decreases for individual


Figure 14. Rare earth SX flow diagram (Mountain Pass, California)

Table 6. U.S. prices for rare earth concentrates and compounds in 2003

Product Purity,* % Quantity, kg Price,† $/kg‡

Concentrates§

Bastnasite concentrate, unleached 58–63 500 3.64 (contained LnO basis)

Bastnasite concentrate, leached 68–73 500 4.08 (contained LnO basis)

Bastnasite concentrate, leached and calcined 85–90 500 5.51 (contained LnO basis)

Lanthanum hydrate 75 minimum 907 5.51 (contained LnO basis)

Lanthanum-lanthanide chloride 46 159 3.20 (contained LnCl basis)

Oxides**

Lanthanum 99.5 1 12.00

Cerium 99.5 1 18.00

Praseodymium 99.5 1 30.00

Neodymium 99.5 1 18.00

Samarium 99.9 1 58.00

Europium 99.99 1 1,120.00

Gadolinium 99.9 1 55.00

Terbium 99.9 1 440.00

Dysprosium 99.5 1 90.00

Holmium 99.9 1 245.00

Erbium 99.9 1 135.00

Thulium 99.9 1 1,950.00

Ytterbium 99.9 1 360.00

Lutetium 99.9 1 2,400.00

* Purity as total lanthanide (Ln) oxides equivalent.† Prices are nominal and subject to change on a daily basis, priced on a contained Ln oxide or Ln chloride basis.‡ Prices for metric-ton quantities or long-term contracts are typically lower.§ Rare earth concentrate prices from Molycorp, Inc., free on board (f.o.b.) Mountain Pass, California.

** REO prices from Hefa Rare Earth Canada Co. Ltd., 1 kg quantity, f.o.b. Vancouver, Canada.

Purification(Eu)

Alternate Flow

Sm-Eu-GdSolution

SmSolution

Nd-PrSolution

Neodymium Oxide

NdSolution

La-Nd-Pr-Sm-Eu-GdSolution

GdSolution

PrSolution

Precipitation, Filter,and Calcine

Praseodymium Oxide


La(NO3)3

HCl

LaCl3

HNO3

Sm-Eu-GdSolution

EuSolution

Sm-GdSolution

Alte

rnat

e Fl

ow

Lanthanide ChlorideSolution without Ce

Filter SX Mixers SX LoadingMixer–Settlers

SX StrippingMixer–Settlers

EuRedox Separation

La-Nd-PrSolution

SX(La-Nd-Pr Separation)

SX(Nd-Pr Separation)

La-Ln Concentrate

Gadolinium Oxide


Samarium Oxide


Europium Oxide


SX(Sm-Gd Separation)


Table 7. Comparative U.S. prices for rare earth metals and oxides, US$/kg

Year

Oxides* 1963† 1973‡ 1983§ 1993** 2003††

Lanthanum 12 10 19 19 12Cerium 17 11 20 23 18Praseodymium 88 71 130 37 30Neodymium 66 26 80 88 18Samarium 99 66 130 66 58Europium 1,411 992 1,900 992 1,120Gadolinium 176 99 140 121 55Terbium 860 606 1,200 827 440Dysprosium 187 88 110 132 90Holmium 254 265 650 485 245Erbium 187 99 200 143 135Thulium 2,756 2,205 3,400 2,750 1,950Ytterbium 331 187 225 220 360Lutetium 7,661 4,409 5,200 5,500 2,400Yttrium 119 66 94 80 52

Year

Metals‡‡ 1963§§ 1973*** 1983††† 1993‡‡‡ 2003§§§

Lanthanum 309 88 125 150 25Cerium 304 88 125 350 30Praseodymium 386 353 310 540 70Neodymium 386 220 260 340 30Samarium 397 298 330 300 80Europium 3,307 5,952 7,500 7,600 1,600Gadolinium 463 463 485 500 78Terbium 2,315 1,543 2,800 2,800 630Dysprosium 661 265 300 500 120Holmium 661 606 1,600 1,400 350Erbium 664 309 650 725 180Thulium 8,378 5,291 8,000 6,500 3,000Ytterbium 1,047 507 875 1,200 484Lutetium 9,370 12,125 14,200 13,000 4,000Yttrium 717 309 430 340 96

* REO prices from U.S., French, and Chinese producers.† Cerium and lutetium prices from American Potash & Chemical Corp.,

2–99-lb quantity, f.o.b. Chicago, Illinois; all other REO prices from Michigan Chemical Corp., 2–99-lb quantity, 99.9% purity, f.o.b., St. Louis, Missouri.

‡ REO prices from Research Chemicals, Inc., 2–49-lb quantity, 99.9% purity, f.o.b., Phoenix, Arizona.

§ REO prices from Research Chemicals, Inc., 1–5-kg quantity, 99.9% purity, f.o.b., Phoenix, Arizona.

** Holmium, thulium, ytterbium, lutetium, and yttrium oxide prices from Rhône-Poulenc Basic Chemicals Co., 1-kg quantity, f.o.b. Phoenix, Arizona; all other oxide prices from Molycorp, Inc., 25–300-lb quantity, f.o.b. Mountain Pass, California.

†† REO prices from Hefa Rare Earth Canada Co. Ltd., 1-kg quantity, 99.5%–99.99% purity, f.o.b. Vancouver, Canada.

‡‡ Rare earth metal prices from U.S., French, and Chinese producers.§§ Rare earth metal prices from American Potash & Chemical Corp., 1–4-lb

quantity, f.o.b., Chicago, Illinois.***Rare earth metal prices from Research Chemicals, Inc., converted from

2–10-lb ingot price, f.o.b., Phoenix, Arizona.†††Rare earth metal prices from Research Chemicals, Inc., 1–5-kg quantity,

ingot price, f.o.b., Phoenix, Arizona.‡‡‡Rare earth metal prices from Rhône-Poulenc Basic Chemicals Co., 1–5-kg

quantity, ingot price, f.o.b., Phoenix, Arizona.§§§Rare earth metal prices from Hefa Rare Earth Canada Co. Ltd., 1-kg quan-

tity, 99.5%–99.99% purity, f.o.b. Vancouver, Canada.

REE commodities. Samarium oxide, for which demand oftenexceeded supply during the late 1980s, sold for $187/kg in 1988but currently is available at $58/kg because neodymium-iron-boron magnets have supplanted much of its market share.Europium oxide also has decreased in price from about $2,000/kgin 1980 to $1,120/kg in 2003. REE prices are generally lower formetric-ton quantities or long-term contracts.

Yttrium raw materials varied greatly in price during the1980s, selling for as much as $50/kg for contained Y2O3 in a 60%concentrate in the mid-1980s, retreating to $32/kg by 1989. In the1990s, yttrium-enriched ion-adsorption lateritic clays from south-ern China began supplying increasing amounts of low-cost yttriumoxide, significantly reducing demand for other yttrium concentratesderived from monazite and xenotime. The availability of low-costyttrium oxide ($7 to $10/kg in metric-ton quantities in 2003) hasessentially eliminated the competition in yttrium concentrate.

Rare Earth Uses and MarketsREEs are used in a wide variety of applications (Table 8). Compris-ing one of the widest ranges of consumer products of any elementgroup, their usage is believed to be a significant economic indicator.Traditional REE applications in lighter flints, carbon arc lighting,and iron and steel additives have declined from 75% of the overallinternational REE market in the 1950s to about 20% in the 2000s.

Table 8. REE uses and applications by industry

Automotive Catalysts for pollution control; catalytic converter catalyst substrate; rechargeable batteries; fuel cells; colored plastics

Ceramics Oxygen sensors; structural ceramics for bearings; jet engine coatings; investment molds; refractories; pigments

Chemicals Oil refinery fluid cracking catalysts; pharmaceuticals; water treatment; catalysts; moisture control, dryers, and detection

Defense Lasers; missile guidance and control; visual displays; radar; electronic countermeasures; communication; shielding

Electronics Capacitors; cathodes; electrodes; semiconductors; thermistors; traveling wave tubes (TWTs); radio frequency circulators and toroids; yttrium iron garnet (YIG) ferrites

Glass Polishing compounds; decolorizing; colorizing; increase refraction; decrease dispersion; radiation stabilization; absorber

Illumination Trichromatic fluorescent lamps; mercury lamps; carbon arc lamps; gas mantles; auto headlamps; long-glow phosphors

Magnets Speakers and headphones; linear motors; antilock braking systems; tape and disk drives; gauges; electric motors; pumps; ignition

Magnetostrictive Sonar systems; precise actuators; precision positioning; vibratory screens; speakers; ultrasonics to kill bacteria

Medical Contrast agents; magnetic resonance imaging (MRI); positron emission tomography (PET); radioisotope tracers and emitters

Metallurgy Alloying agents in aluminum, magnesium, iron, nickel, and steel alloys; superalloys; pyrophoric alloys; lighter flints; armaments

Phosphors Cathode-ray tubes (CRTs); fluorescent lighting; radar and cockpit displays; x-ray intensifying screens; temperature sensors

Other Simulated gemstones; textiles; magnetic refrigeration; hydrogen fuel storage; lubrication; photography; nuclear uses


Uses in glass polishing and ceramics are the largest market for REEcommodities by weight (Figure 15). Substantial amounts of ceriumconcentrates and cerium oxide are used in glass-polishing applica-tions. The automotive catalyst industry also consumes significantREEs by weight, using cerium carbonate and cerium oxide in thecatalyst substrate and as a component of the converter’s oxidizingcatalyst system. During the 1980s, REE production increasesresulted from the rising use of cerium in automotive catalytic con-verters (O’Driscoll 1988). The use of lanthanum-rich mixed REEcompounds in fluid cracking catalysts (FCCs) in petroleum refiningalso has been a major market (Figure 15). Although this applicationinitially declined substantially in the mid-1980s when legislationmandated the use of unleaded fuels, the yield of gasoline and otherlight-fraction products was so reduced that consumption of REE-containing FCCs were restored. Phosphors, electronics, and lasercrystals provide relatively minor markets by weight, although mostare high value.

Although it only comprises 3% of REE production by weight,the phosphor market is probably the most important in terms of dol-lar value. Beginning in the 1960s, color television manufacturersspurred demand for high-value REE phosphors based on yttriumand europium; later, gadolinium and terbium compounds wereadded as phosphors. More recently, their use in trichromatic phos-phors in fluorescent lighting resulted in increased production.

The introduction of high-strength samarium-cobalt permanentmagnets developed by Karl Strnat in the 1970s, particularly for usein small electric motors and headphones, created a dynamic newREE market. Although higher-strength, lower-cost neodymium-iron-boron permanent magnets are now substituted for the moreexpensive samarium-cobalt magnets in most applications, samar-ium-cobalt magnets are still used in applications requiring highertemperatures.

Proposed new uses have increased awareness of potentialprofits in REE raw materials. Superconducting substances contain-ing lanthanum or yttrium have been in the spotlight since 1986.Volume markets for these substances, however, await further tech-nological breakthroughs. New methods have been devised for REEsused in hydrogen-based refrigeration and energy storage. Marketexpansion for yttrium has provided an incentive for increased pro-duction from HREE deposits. In addition to its use in television,computer, and lighting phosphors, yttrium has found increased usein stabilized-zirconia ceramics applied in structural parts, bearings,

Figure 15. Rare earth markets by weight

Petroleum RefiningCatalysts

28%

PermanentMagnets

3%

MetallurgicalAdditives and

Alloys19%

Glass Polishing and Ceramics

30%

AutomotiveCatalytic

Converters14%

Rare EarthPhosphors

3%Miscellaneous3%

oxygen sensors, simulated gemstones, golf clubs and shoe cleats,shirt buttons, and coatings in the exhaust section of jet engines.

Although worldwide demand for REEs has had a generalupward trend for many decades, REEs’ economics are as complexas their chemistry. Because individual REEs occur together in mostdeposits in fixed ratios, the economic riddle is how to balance pro-duction against market demand. To accomplish this, each REEmust be assigned a factor based on its natural abundance in the oremineral in order to develop marketing equations, such as determi-nation of price. Complexity of these equations are enhanced byvariations in separation and purification costs between the REEs. Inaddition, markets for individual REEs may change rapidly becauseof technological changes and other factors. Marketing variabilitymay also lead to stockpiling of unmarketable REE fractions andproducts by producers.

The principal source of REE ores is China, with smalleramounts available from India, Kazakhstan, Kyrgyzstan, Malaysia,Russia, Thailand, and the United States. Refined rare earth prod-ucts, including compounds, metals, and alloys, are primarily pro-duced in China and France.

U.S. imports were dominated by Chinese material in 2003,with 75% of domestic imports, by weight. France, with 10%, wasthe next largest source of U.S. imports (down from 30% in 2000),and Japan was third, with 4%. France and Japan, however, do notmine REEs and source the majority of their supplies from China.Therefore, the amount of Chinese REEs that the United States actu-ally imported in 2003 was probably 89%. For the rare earth tradecategories of domestic imports, the percentage of shipments sup-plied by China, by weight, were as follows:

• Cerium compounds: 84%

• Yttrium compounds: 63%

• Individual REE compounds (except cerium): 64%

• Mixtures of REOs (except cerium): 97%

• REE metals (intermixed or interalloyed): 73%

• Mixtures of REE chlorides: 94%

China was the leading source of materials for all rare earth importcategories except ferrocerium and other pyrophoric alloys, a smallvolume trade category supplied mainly from France, at 82% of theimport market.

In the mid-1980s, U.S. imports exhibited a very different dis-tribution, at lower quantities because of high REE production fromMountain Pass. The leading supplier of domestic imports in 1985in each category by percentage of shipments by weight, were asfollows:

• Cerium compounds (except oxide): France, 89%

• Cerium oxide: France, 98%

• Individual REOs (excluding cerium): France, 84%

• REE alloys (including mischmetal): Brazil, 87%

• REE metals: former U.S.S.R., 65%

• Other REE metals: Federal Republic of Germany, 21%

• Ferrocerium and other pyrophoric alloys: France, 59%

In 1985, the majority of rare earth imports came from France,at 48%, with China supplying only about 1%.

Domestic production trended upward from the 1960s, whenbastnasite production from Mountain Pass became commerciallyavailable, to the 1990s. Concentrate production at Mountain Passceased during 2001, when the mine and mill were put on standbystatus. Stocks of bastnasite concentrates and some REE intermedi-ate and purified compounds are available from stockpiles, however.


Table 9. Harmonized Tariff Schedule for rare earth commodities imported into the United States in 2004

Heading/ Subheading Codes Article Description

Normal TradeRelations

Non-normal Trade Relations Special Program Codes*

2530.90.8050 Mineral substances not elsewhere specified or included; rare earth metal ores, e.g., xenotime (complex phosphate) and gadolinite, ytterbite, and cerite (complex silicates)

Free 0.3¢/kg

2612.20.0000 Thorium ores and concentrates; rare earth metal ores, e.g., monazite (thorium and rare earth phosphate)

Free Free

2805.30.0000 Rare earth metals, including scandium and yttrium, whether intermixed or interalloyed, ad valorem

5.0% ad valorem 31.3% ad valorem Free A+, CA, CL, D, E, IL J, JO, MX, 3.7% (SG)

2846.10.0000 Cerium compounds, including oxides, carbonates, hydroxides, nitrates, sulfates, chlorides, oxalates, and so forth

5.5% ad valorem 35% ad valorem Free A, CA, CL, E, IL, J, JO, MX, 4.1% (SG)

2846.90.2010 Mixtures of rare earth oxides, including yttrium and scandium, other than cerium oxides

Free 25% ad valorem

2846.90.2050 Mixtures of rare earth chlorides, except cerium chloride Free 25% ad valorem

2846.90.4000 Yttrium-bearing materials and compounds containing by weight >19% but <85% yttrium oxide equivalent

Free 25% ad valorem

2846.90.8000 Rare earth compounds, including individual REOs, rare earth hydroxide, rare earth nitrate, and other compounds (excludes cerium compounds, mixtures of REOs, and mixtures of rare earth chlorides)

3.7% ad valorem 25% ad valorem Free A, CA, CL, E, IL, J, JO, K, MX, SG

2844.40.0020 Other radioactive elements, isotopes, and compounds (includes promethium, gadolinium 157, and so forth)

Free Free

3606.90.3000 Ferrocerium and other pyrophoric alloys 5.9% ad valorem 56.7% ad valorem Free A+, CA, CL, D, E, IL, J, JO, MX, 4.4% (SG)

7202.99.5040 Other ferrosilicon alloys, containing by weight not more than 55% silicon, other (e.g., rare earth silicide)

Free 4.4¢/kg on silicon content

* Products eligible for special tariff treatment:A = Generalized System of Preferences (GSP) (duty-free treatment).

A+ = Only imports from least-developed beneficiary developing countries eligible for GSP under that subheading (duty-free treatment).

CA = North American Free Trade Agreement (NAFTA) for Canada (duty-free treatment).

CL = Chile Special Rate.D = Africa Growth and Opportunity Act (duty-free treatment).

E = Caribbean Basin Initiative (CBI).IL = Israel Special Rate (duty-free treatment).J = Andean Trade Preference Act (ATPA).

JO = Jordan Special Rate.K = Agreement on Trade in Pharmaceutical Products (duty-free treatment).

MX = NAFTA for Mexico (duty-free treatment).SG = Singapore Special Rate.

Although REE raw material production is dominated by onecountry, worldwide REE commodity markets are complex becauseof the variety of ores and commodities. Also, REE separation facili-ties are located in various countries that import all of their plant feedmaterial, which means that high-purity REE products may beshipped back to the countries that originally produced the REE rawmaterial. France and Japan, which have no internal REE raw mate-rial sources, are major producers of processed REE commodities,exporting approximately $20 million and $15 million worth, respec-tively, of REE compounds and metals to the United States in 2003.Separated REE production in France is by a single company, RhodiaElectronics & Catalysis. Originally, it utilized monazite and xeno-time from Australia and the United States; however, restrictions onthe disposal of thorium waste forced Rhodia to shift to intermediateREE thorium-free compounds, including rare earth chlorides andnitrates. Sixteen European countries exported REE commodities tothe United States in 2003: Austria, Belgium, Estonia, France, Ger-many, Hungary, Ireland, Italy, Liechtenstein, the Netherlands, Nor-way, Russia, Spain, Sweden, Switzerland, and the United Kingdom.At least 20 companies in Japan produce REE commodities that areused by that country’s steel, auto, and electronics industries. Japa-nese REE producers have used increasing amounts of Chinese rawand purified REE materials for their industry.

GOVERNMENTAL CONSIDERATIONSTariff Rates, Depletion Provisions, and Government PoliciesThe classification of China as a most favored nation in 1978 by theUnited States had a strong impact on REE markets, and the U.S.

government has shown little interest in protecting domestic produc-ers. U.S. tariff rates are shown in Table 9.

While keeping the value of its currency low, the Chinese gov-ernment has aggressively promoted its REE industry. Peking Univer-sity is reported to have a research staff of 1,400 scientists working onall facets of REE production and marketing (D. Witham, personalcommunication). Development of Chinese REE deposits does notseem to have been impeded by normal market considerations, leadingto significant overproduction of REEs in recent years. Many ofChina’s REE mines, processing facilities, and manufacturing plantsare said to be underfunded (O’Driscoll 2003). In 2003, the Chinesegovernment announced that it would assert control over its REEindustry, setting up two oversight groups: one centered on the BayanObo–Baotou operations in the north and one that will be developedmainly around the lateritic (ionic clay) REE deposits in southernChina.

Environmental ConcernsIn the United States, REE mines and processing plants arerequired to meet the same environmental quality guidelines asother mining and chemical facilities. The U.S. EnvironmentalProtection Agency (EPA) is responsible for setting air-, water-,and soil-quality standards for such operations nationwide, butstandards for some individual states are more stringent. Reclama-tion requirements for mining are under state authority and differfrom state to state; however, reclamation on federally owned landis overseen by agencies such as the U.S. Forest Service and U.S.Bureau of Land Management.


REE ores and commodities, as well as by-products and wastematerials from REE processing, are naturally radioactive, mostlybecause of contained thorium. The thorium content of monazitetypically ranges from 4% to 10% ThO2, whereas thorium com-prises less than 1,000 ppm in bastnasite from Mountain Pass (Mari-ano 1989a) and may be even lower in Chinese REE concentrate.Handling, shipping, storage, and disposal of such materials are con-trolled by federal and state agencies. Less than 10 t of thorium com-modities, all from imported or stockpiled material, are consumed inthe United States annually (Hedrick 2004b). Therefore, thoriumproduced during REE processing presents an expensive disposalproblem for the industry.

The Rhodia plant in Freeport, Texas, extracted REEs fromlargely imported monazite until 1992, when the feedstock wasswitched to Chinese REE concentrates. Production from this opera-tion has declined substantially in recent years. The company’s LaRochelle, France, plant is now the world’s only significant REEseparation plant outside of China, and it imports most of its REEsas carbonates, chlorides, and nitrates from China (Chegwidden andKingsnorth 2002).

Although the bastnasite produced by Molycorp, Inc., atMountain Pass has relatively low thorium, chemical processing wassuspended by the California EPA in 1998, mainly because of theradioactive element content of waste involved in a spill incident.Between 1998 and the present, Molycorp has sold REE commodi-ties from stockpile and has purchased, tested, and resold importedREE commodities. The company plans to resume production atMountain Pass and its environmental impact statement is currentlyunder review.

Environmental concerns centered on the disposal of radioac-tive waste have precluded the development of REE extractionplants in Australia, where environmental policies are mainly understate control. A processing facility in Western Australia designed toproduce REEs from 15,000 t of monazite annually, originally slatedto begin production in 1989, was not constructed because of lack ofapproval from the Environmental Protection Authority of WesternAustralia. Groups concerned about radioactive contamination of theenvironment have become active worldwide, as shown by protestsof plans to begin processing REEs in Malaysia in the late 1980s.Malaysian production of xenotime ore ceased in 1994 because ofenvironmental regulatory compliance problems, along with the col-lapse of market prices for xenotime.

On the other hand, consumption of REEs in automotive andindustrial emission catalysts is directly dependent on environmentalregulations for air pollution control. Use and composition of REE-containing refinery FCCs are functions of both the feedstock andthe fuel quality required, both of which also relate back to air pollu-tion regulations.

FUTURE TRENDSThe future use of REEs is expected to increase, especially in auto-motive pollution catalysts, FCCs, and permanent magnets. Rareearth applications are expected to need greater amounts of higher-purity mixed and separated products, requiring specificity in theirunique properties. Strong demand for cerium, lanthanum, andneodymium is expected to continue throughout the decade as auto-motive and electronic usage expands with population growth.Future growth also is forecast for REEs used in lasers, fiber optics,and medical applications that include MRI contrast agents, PETscintillation detectors, medical isotopes, and dental and surgicallasers. On the other hand, demand is expected to decline over thenext decade for REE phosphors used in CRTs, as flat-panel displayuse increases. A slight decline also is forecast for rechargeable

nickel-hydride batteries, as lithium ion batteries gain wider use inportable electronics. Long-term growth is forecast for REEs uti-lized in magnetic refrigeration alloys and other technologies thatexhibit energy efficiency and cost benefits.

World reserves are sufficient to meet forecast world demandwell into the 21st century. Several world-class REE deposits inAustralia and China (Mount Weld and Maoniuping, respectively)have yet to be fully developed and contain substantial reserves. Pro-duction may resume at idle deposits that produced in the past, suchas Mountain Pass, United States, and Steenkampskraal, SouthAfrica. Other substantial reserves are contained in heavy-mineralsands deposits worldwide. As demand continues to increase, futuredevelopment of new REE deposits is likely.

U.S. companies have shifted away from using radioactiveREE ores, which has negatively affected monazite-containing min-eral-sands operations worldwide. The cost to dispose of radioactivewaste products in the United States is expected to continue toincrease, severely limiting domestic use of low-cost monazite andother thorium-bearing REE ores. Worldwide demand for monaziteis expected to increase in the long term, however, because of itsabundant supply and its recovery as a low-cost by-product.

World REE markets are expected to continue to be very com-petitive in response to China’s large resources, competitive prices,low-cost wages, inexpensive utilities, and minimal environmentaland permitting requirements. China is expected to remain theworld’s principal rare earth supplier. In addition, cconomic growthin several developing countries will provide new and potentiallylarge markets in Southeast Asia and Eastern Europe.

The long-term outlook is for an increasingly competitive anddiverse group of REE suppliers. As research and technology con-tinue to advance the knowledge of REEs and their interactions withother elements, the economic base of the REE industry is expectedto continue to grow. New REE applications are expected to con-tinue to be discovered and developed.

BIBLIOGRAPHYAdams, J.W. 1965. Rare earths. Pages 234–237 in Mineral and

Water Resources of New Mexico. Bulletin 87. Socorro: NewMexico Bureau of Mines and Mineral Resources.

Altschuler, Z.S., S. Berman, and F. Cuttitta. 1967. Rare earths inphosphorite—geochemistry and potential recovery. ProfessionalPaper 575-B. Reston, VA: U.S. Geological Survey (USGS).

Anderson, J.L. 1983. Proterozoic anorogenic granite plutonism ofNorth America. Pages 133–154 in Proterozoic Geology:Selected Papers from an International Proterozoic Symposium.Edited by L.G. Medaris, C.W. Byers, D.M. Mickelson, and W.C.Shanks. Memoir 161. Boulder, CO: Geological Society ofAmerica.

Andreoli, M.A.G., C.B. Smith, M. Watkeys, J.M. Moore, L.D.Ashwal, and R.J. Hart. 1994. The geology of theSteenkampskraal monazite deposit, South Africa: Implicationsfor REE-Th-Cu mineralization in charnockite-granuliteterranes. Economic Geology 89:994–1016.

Anon. 1980. Mineral Powder Diffraction File. Swathmore, PA:Joint Committee on Powder Diffraction Standards.

———. 1984. Carbonatitic Complexes of Brazil: Geology. SaoPaulo: Companhia Brasileira de Metalurgia e Mineraçao.

———. 1987. Mineral sands, the big scale time. IndustrialMinerals 231:47–61.

———. 2003. Nolans Bore rare earths/phosphate. IndustrialMinerals 424:71.

Argall, G.O. 1980. Three iron ore bodies of Bayan Obo. WorldMining (January): 38–41.


Armbrustmacher, T.J. 1980. Abundance and distribution of thoriumin the carbonatite stock at Iron Hill, Powderhorn District,Gunnison County, Colorado. Professional Paper 1049-B.Reston, VA: USGS.

———. 1988. Geology and resources of thorium and associatedelements in the Wet Mountains Area, Fremont and CusterCounties, Colorado. Professional Paper 1049-F. Reston, VA:USGS.

Australian Zirconia Ltd. 2000. Dubbo zirconia project. Companyliterature. Dubbo, New South Wales: Australian Zirconia Ltd.

Bai, G., and Z. Yuan. 1985. On the REE-rich carbonatites. Pages126–136 in New Frontiers in REE Science and Applications.Edited by X. Quangxian. Beijing, China: Science Press.

Barker, J.C., and C.L. Mardock. 1988. Lithophile metal, REE-Y-Nbdeposits on Southern Prince of Wales Island, Alaska. Pages139–158 in Process Mineralogy VIII. Edited by D.J.T. Carsonand A.H. Vassiliou. Warrendale, PA: Minerals, Metals andMaterials Society.

Borrowman, S.R., and J.D. Rosenbaum. 1962. Recovery ofThorium from a Wyoming Ore. Report of Investigation 5917.Washington, DC: U.S. Bureau of Mines (USBM).

Campbell, L.S., and P. Henderson. 1997. Apatite paragenesis in theBayan Obo REE-Nb-Fe ore deposit, Inner Mongolia, China.Lithos 42:89–103.

Castor, S.B. 1986. Mountain Pass Bastnasite—Some New Data.Technical Report. Mountain Pass, CA: Molycorp.

———. 1991. Rare earth deposits in the Southern Great Basin.Pages 523–528 in Geology and Ore Deposits of the Great Basin.Edited by G.L. Raines, R.E. Lisle, R.W. Schafer, and W.H.Wilkinson. Reno: Geological Society of Nevada.

———. 1993. Rare earth deposits and Proterozoic anorogenicmagmatism. Pages 331–340 in Proceedings of the 8thQuadrennial IAGOD Symposium, 1990, Ottawa, Canada. Editedby Y.T. Maurice. Stuttgart, Germany: Schweizerbart’sche Ver-lagsbuchhandlung.

———. 1994. Rare earth minerals. Pages 827–839 in IndustrialMinerals and Rocks. 6th edition. Edited by D.D. Carr. Littleton,CO: SME.

Castor, S.B., and G.W. Nason. 2004. Mountain Pass rare earthdeposit, California. Pages 68–81 in Proceedings of the 39thForum on the Geology of Industrial Minerals. Edited by. S.B.Castor. Special Publication 33. Reno: Nevada Bureau of Minesand Geology.

Chao, E.C.T., et al. 1997. The Sedimentary Carbonate-HostedGiant Bayan Obo REE-Fe-Nb Ore Deposit of Inner Mongolia,China; a Cornerstone Example for Giant Polymetallic OreDeposits of Hydrothermal Origin. Bulletin 2143. Reston, VA:USGS.

Chegwidden, J., and D.J. Kingsnorth. 2002. Rare earths supplyand demand, a European market focus. Industrial Minerals415:52–61.

Costa, M.L., L.R. Fanesca, R.S. Angelica, V.P. Lemos, and R.L.Lemos. 1991. Geochemical exploration of the Maicuru alkaline-ultramafic-carbonatite complex, northern Brazil. Journal ofGeochemical Exploration 40:193–204.

Crowley, F.A. 1960. Columbium—Rare Earth Deposits, SouthernRavalli County, Montana. Bulletin 18. Montana Bureau ofMines.

Currie, K.L. 1976. The Alkaline Rocks of Canada. Bulletin 239.Geological Survey of Canada.

———. 1985. An unusual peralkaline granite near Lac Brisson,Quebec–Labrador. Pages 73–80 in Current Research, Part A.Paper 85-1A. Geological Survey of Canada.

Deans, T. 1966. Economic geology of African carbonatites. Pages385–413 in Carbonatites. Edited by O.F. Tuttle and J. Gittins.New York: Wiley Interscience.

DeWitt, E., L.M. Kwak, and R.E. Zartman. 1987. U-Th-Pb and Ar/Ar dating of the Mountain Pass carbonatite and alkalic igneousrocks, S.E. Cal. Geological Society of America Abstracts withPrograms 19(7):642.

Drew, L.J., Q. Meng, and W. Sun. 1990. The Bayan Obo iron-rareearth-niobium deposits, Inner Mongolia, China. Lithos 26:43–65.

Drysdall, A.R., N.J. Jackson, C.R. Ramsay, C.J. Douch, and D.Hackett. 1984. Rare element mineralization related toPrecambrian alkali granites in the Arabian Shield. EconomicGeology 79:1366–1377.

Duncan, R.K., and G.C. Willett. 1990. Mount Weld carbonatite.Pages 591–597 in Geology of the Mineral Deposits of Australiaand Papua New Guinea. Edited by F.E. Hughes. Melbourne:The Australasian Institute of Mining and Metallurgy.

Dunn, C.E. 1995. Biogeochemical prospecting for metals. Pages371–524 in Biological Systems in Mineral Exploration andProcessing. Edited by R.R. Brooks, C.E. Dunn, and G.E.M.Hall. New York: Ellis Horwood.

Emsley, J. 2001. Nature’s Building Blocks: An A–Z Guide to theElements. Oxford, England: Oxford University Press.

Evans, J.R. 1964. Xenotime Mineralization in the Southern MusicValley Area, Riverside County, California. Special Report 79.California Division of Mines and Geology.

Ferguson, J. 1970. The differentiation of agpaitic magmas: TheIlimaussaq intrusion, south Greenland. Canadian Mineralogist10:335–49.

Frondel, C. 1958. Systematic Mineralogy of Uranium and Thorium.Bulletin 1064. Reston, VA: USGS.

Fryer, B.J., and A.D. Edgar. 1977. Significance of rare earth distri-butions in coexisting minerals of peralkaline undersaturatedrocks. Contributions to Mineralogy and Petrology 61:35–48.

Garson, M.S. 1966. Carbonatites in Malawi. Pages 33–72 inCarbonatites. Edited by O.F. Tuttle and J. Gittins. New York:Wiley Interscience.

Geological Survey of Kyrgyzstan. 2004. Rare earth elements,Kutessai II deposit. Investment Project. http://www.kgs.bishkek.gov.kg/invest_eng.htm. Accessed July 2004.

Gerasimovsky, V.I., V.P. Volkov, L.N. Kogarko, and A.I. Polyakov.1974. Alkaline provinces: Kola Peninsula. Pages 206–221 inThe Alkaline Rocks. Edited by H. Sorensen. London: JohnWiley & Sons.

Griffiths, S.J. 1984. Review of placer deposits in Australia. Pages265–307 in Placer Exploration and Mining, A Short Course.Edited by C. McKay and P. Mousset-Jones. Reno: University ofNevada Division of Continuing Education.

Gulbrandsen, R.A. 1966. Chemical compositions of phosphoritesof the Phosphoria Formation. Geochemica et CosmochimicaActa 30:769–778.

Gunn, P.J., and M.C. Dentith. 1997. Magnetic responses associatedwith mineral deposits. AGSO Journal of Australian Geology andGeophysics 17(2):145–158.

Hatzl, T., G. Morteani, A. Fuganti, P. Kienast, and D. Blamart.1990. Petrology and geochemistry of the bastnaesite-fluorite-barite mineralization near Kizilçaören, Province Eskisehir.Pages 83–84 in International Earth Sciences Congress onAegean Regions, 1–6 October 1990, Abstracts. Izmir, Turkey:IESCA.

Haxel, G.B., J.B. Hedrick, and G.J. Orris. 2002. Rare EarthElements—Critical Resources for High Technology. Fact sheet087-02. Reston, VA: USGS.


Hedrick, J.B. 1985. Rare earth elements and yttrium. Pages 647–664 in Mineral Facts and Problems. Bulletin 675. Washington,DC: USBM.

———. 1991a. Rare earth metals. Pages 128–129 in MineralCommodity Summaries. Washington, DC: USBM.

———. 1991b. Thorium. Pages 128–129 in Mineral CommoditySummaries. Washington, DC: USBM.

———. 1992. Rare Earth Minerals and Metals, 1990. Washington,DC: USBM.

———. 2003. Rare earths. Pages 60.1–60.9 in Minerals Yearbook2003. Reston, VA: USGS.

———. 2004a. Rare earths. Pages 132–133 in Mineral CommoditySummaries 2003. Reston, VA: USGS.

———. 2004b. Thorium. Pages 172–173 in Mineral CommoditySummaries 2003. Reston, VA: USGS.

Hedrick, J.B., and S.P. Sinha. 1994. Cerium-based polishingcompounds—discovery to manufacture. Journal of Alloys andCompounds 207–208:377–382.

Hedrick, J.B., S.P. Sinha, and V.D. Kosynkin. 1997. Loparite, a rareearth ore (Ce, Na, Sr, Ca)(Ti, Nb, Ta, Fe+3)O3. Journal of Alloysand Compounds 250: 467–470.

Hedrick, J.B., and D.A. Templeton. 1991. Rare Earth Minerals andMetals, 1989. Washington, DC: USBM.

Heinrich, E.W., and A.A. Levinson. 1961. Carbonatite rare earthdeposits, Ravalli County, Montana. American Mineralogist46:1424–1447.

Henderson, P. 1984. General geochemical properties and abun-dances of the rare earth elements. Pages 1–32 in Rare Earth Ele-ment Geochemistry. Edited by P. Henderson. New York:Elsevier.

Hitzman, M.W., N. Oreskes, and M.T. Einaudi. 1992. Geologicalcharacteristics and tectonic setting of Proterozoic iron oxide(Cu-U-Au-REE) deposits. Precambrian Research 58:241–287.

Issa Filho, A., P.R.A. Lima, and O.M. Souza. 1984. Aspects of thegeology of the Barreiro carbonatitic complex, Araxá, MG,Brazil. Pages 21–44 in Carbonatitic Complexes of Brazil:Geology. Sao Paulo, Brazil: Companhia Brasileira deMetalurgia e Mineraçao.

Jackson, W.D., and G. Christiansen. 1993. International strategicminerals inventory summary report—rare earth oxides. Circular930-N. Reston, VA: USGS.

Jones, A.P., F. Wall, and C.T. Williams. 1996. Rare Earth Minerals:Chemistry, Origin and Ore Deposits. Mineralogical SocietySeries 7. London: Chapman and Hall.

Kaczmarek, J. 1980. Rare earths: Discovery and commercialseparations. Address to the American Chemical Society.Monmouth Junction, NJ: Rhône-Poulenc.

Kapustin, Y.L. 1980. Mineralogy of Carbonatites. New Delhi:Amerind Publishing.

Kelley, V.C. 1968. Geology of alkali Precambrian rocks at PajaritoMountain, Otero County, New Mexico. Geological Society ofAmerica Bulletin 79:1565–1572.

Kisvarsanyi, E.B., P.K. Sims, and G. Kisvarsanyi. 1989.Proterozoic ore-forming systems: Examples of ironmetallogenesis from the United States midcontinent. GeologicalSociety of America Abstracts with Programs 21(6):A34.

Kline, M.H., E.J. Carlson, A.F. Roberston, and R.H. Storch. 1953.Bear Valley Radioactive Mineral Placers, Valley County, Idaho.Report of Mineral Evaluation 3130. Washington, DC: USBM.

Kogarko, L.N., V.A. Kononova, M.P. Orlova, and A.R. Woolley.1995. Alkaline Rocks and Carbonatites of the World, Part Two,Former USSR. London: Chapman and Hall.

Kosynkin, V.D., S.D. Moiseev, C.H. Peterson, and B.V. Nikepelov.1993. Rare earths industry in the Commonwealth ofIndependent States. Journal of Alloys and Compounds 192:118.

Le Bas, M.J., B. Spiro, and X. Yang. 1997. Oxygen, carbon, andstrontium isotope study of the carbonatitic dolomite host of theBayan Obo Fe-Nb-REE deposit, Inner Mongolia, N. China.Mineralogical Magazine 61(4):531–541.

Lissiman, J.C., and R.J. Oxenford. 1975. Eneabba rutile-zircon-ilmenite sand deposit, W.A. Pages 1062–1070 in EconomicGeology of Australia and Papua New Guinea, I: Metals.Monograph 5. Parkville, Victoria: Australasian Institute ofMining and Metallurgy.

Lottermoser, B.G. 1988. Supergene, secondary monazite from theMt. Weld carbonatite laterite, Western Australia. NeuesYahrbuch Mineralogie H. 2:67–70.

———. 1990. Rare earth element mineralisation within the Mt.Weld carbonatite laterite, Western Australia. Lithos 2:151–167.

Lundqvist, T. 1980. The Precambrian of Sweden. Pages 211–271 inGeology of the European Countries. Published in cooperationwith the Comité National Français de Géologie (C.N.F.G.) onthe occasion of the 26th International Geologic Congress. Paris:Dunod Éditeur.

Lurie, J. 1986. Mineralization of the Pilanesberg alkaline complex.Pages 2215–2228 in Mineral Deposits of South Africa, II.Geological Society of South Africa.

Maksimovic, Z.J., and G. Pantó. 1996. Authigenic rare earthminerals in karst-bauxites and karstic nickel deposits. Pages257–279 in Rare Earth Minerals: Chemistry, Origin and OreDeposits. Edited by A.P. Jones, F. Wall, and C.T. Williams.Mineralogical Society Series 7. London: Chapman and Hall.

Mandarino, J.A. 1999. Fleischer’s Glossary of Mineral Species1999. Tucson, AZ: The Mineralogical Record Inc.

Mariano, A.N. 1989a. Economic geology of rare earth elements.Pages 309–337 in Reviews in Mineralogy. Edited by B.R. Lipinand G.A. McKay. Washington, DC: Mineralogical Society ofAmerica.

———. 1989b. Nature of economic mineralization in carbonatitesand related rocks. Pages 149–165 in Carbonatites—Genesis andEvolution. Edited by K. Bell. London: Unwin Hyman.

Mazzi, F., and R. Munno. 1983. Calciobetafite (new member of thepyrochlore group) and related minerals from Campi Klegrei,Italy: Crystal structures of polymignyte and zirkelite:comparison with pyrochlore and zirconolite. AmericanMineralogist 68:262.

McKellar, J.B. 1975. The eastern Australian rutile province. Pages1055–1062 in Economic Geology of Australia and Papua NewGuinea, I: Metals. Monograph 5. Parkville, Victoria:Australasian Institute of Mining and Metallurgy.

McKeown, F.A., and H. Klemic. 1956. Pages 9–23 in Rare Earth–Bearing Apatite at Mineville, Essex County, New York. Bulletin1046-B. Reston, VA: USGS.

Mertie, J.B. 1960. Monazite and related minerals. Pages 623–629in Industrial Minerals and Rocks. 3rd edition. Edited by J.L.Gillson. New York: AIME.

Metz, M.C., D.G. Brookins, P.E. Rosenberg, and R.E. Zartman.1985. Geology and geochemistry of the Snowbird deposit,Mineral County, Montana. Economic Geology 80:394–409.

Mikhaillichenko, A.J., E.B. Mikhlin, and Y.B. Pattrikeev. 1987.Rare earth metals. In Metallugiya (in Russian).

Miller, R.R. 1988. Yttrium (Y) and other rare metals (Be, Nb, REE,Ta, Zr) in Labrador. Pages 229–245 in Current Research (1988).Report 88-1. Newfoundland Department of Mines, MineralDevelopment Division.


Morteani, G., and C. Preinfalk. 1996. REE distribution and REEcarriers in laterites formed on the alkaline complexes of Araxáand Catalão (Brazil). Pages 227–255 in Rare Earth Minerals:Chemistry, Origin and Ore Deposits. Edited by A.P. Jones, F.Wall, and C.T. Williams. Mineralogical Society Series 7.London: Chapman and Hall.

Murata, K.J., and H. Bastron. 1956. Convenient method forrecognizing nonopague cerium-earth minerals. Science123:888–889.

Nakai, S., A. Masuda, H. Shimizu, and L. Qi. 1989. La-Ba datingand Nd and Sr isotope studies on the Baiyun Obo rare earthelement ore deposits, Inner Mongolia, China. EconomicGeology 84:2296–2299.

Neary, C.R., and D.E. Highley. 1984. The economic importance ofthe rare earth elements. Pages 423–466 in Rare Earth ElementGeochemistry. Edited by P. Henderson. Amsterdam: Elsevier.

Nelson, D.R., A.R. Chivas, B.W. Chappell, and M.T. McCullough.1988. Geochemical and isotopic systematics in carbonatites andimplications for the evolution of ocean–island sources.Geochimica et Cosmochimica Acta 52:1–17.

Notholt, A.J.G., D.E. Highley, and T. Deans. 1990. Economicminerals in carbonatites and associated alkaline igneous rocks.Transactions, Institution of Mining and Metallurgy 99:B59–B80.

Nuelle, L.M., W.C. Day, G.B. Sidder, and C.M. Seeger. 1992. PagesA1–A11 in Geology and Mineral Paragenesis of the Pea RidgeIron Ore Mine, Washington County, Missouri; Origin of theRare Earth–Element and Gold-Bearing Breccia Pipes. Bulletin1989. Reston, VA: USGS.

O’Driscoll, M. 1988. Rare earths—enter the dragon. IndustrialMinerals 254:21–54.

———. 2003. Regeneration in autocatalysts. Industrial Minerals424:21–54.

Olson, J.C., and D.C. Hedlund. 1981. Alkalic Rocks and Resourcesof Thorium and Associated Elements in the PowderhornDistrict, Gunnison County, Colorado. Professional Paper 1049-C. Reston, VA: USGS.

Olson, J.C., L.C. Pray, D.R. Shawe, and W.N. Sharp. 1954. RareEarth Mineral Deposits of the Mountain Pass District, SanBernardino County, California. Professional Paper 261. Reston,VA: USGS.

Oreskes, N., and M.T. Einaudi. 1990. Origin of rare earth element-enriched hematite breccias at the Olympic Dam Cu-U-Au-Agdeposit, Roxby Downs, South Australia. Economic Geology85:1–28.

Orris, G.J., and R.I. Grauch. 2002. Rare Earth Element Mines,Deposits, and Occurrences. Open-File Report 02-189. Reston,VA: USGS.

Overstreet, W.C. 1967. The Geologic Occurrence of Monazite.Professional Paper 530. Reston, VA: USGS.

Parak, T. 1973. Rare earths in the apatite iron ores of Lapplandtogether with some data about the Sr, Th, and U content of theseores. Economic Geology 68:210–221.

Philpotts, J.A., M. Tatsumoto, K. Wang, and P. Fan. 1989.Petrography, chemistry, age, and origin of the REE-iron depositat Bayan Obo, China, and implications of Proterozoic iron oresin earth evolution. Pages 53–55 in USGS Research on MineralResources, 1989; Program and Abstracts. Circular 1035.Reston, VA: USGS.

Powell, J.L., P.M. Hurley, and H.W. Fairbairn. 1966. The strontiumisotopic composition and origin of carbonatites. Pages 365–378in Carbonatites. Edited by O.F. Tuttle and J. Gittins. New York:John Wiley & Sons.

Premoli, C. 1989. Vietnamese metallic mineral resources. Engi-neering and Mining Journal 190(12):31–35.

Pu, G., H. Chen, and Z. Yang. 2001. REE metallogenic history andfeatures in Panxi region, Sichuan. Pages 1–5 in Proceedings ofthe 4th International Conference on Rare Earth Developmentand Application. Edited by Y. Zongsen, Y. Chunhua, X.Guangyao, N. Jingkao, and C. Zhanheng. Beijing, China:Metallurgical Industry Press.

Qiu, Y., Z. Wang, and Z. Zhao. 1983. Preliminary study of REEiron formations in China. Geochemistry 2:185–200.

Ren, X. 1985. A brief account of rare earth mineralization in China.Pages 39–41 in New Frontiers in REE Science and Applications.Edited by Q. Xu. Beijing, China: Science Press.

Richardson, D.G., and T.C. Birkett. 1996. Peralkaline rock-associated rare metals. Pages 523–540 in Geology of CanadianMineral Deposit Types. Edited by O.R. Eckstrand, W.D.Sinclair, and R.I. Thorpe. Geology of Canada 8. GeologicalSurvey of Canada.

Robertson, J.A. 1981. The Blind River uranium deposits: The oresand their setting. Pages U1–U23 in Genesis of U- and Au-Bearing Quart–Pebble Conglomerates. Edited by F.C.Armstrong. Professional Paper 1161. Reston, VA: USGS.

Roeder, P.K., D. MacArthur, X. Ma, G.R. Palmer, and J.N.Mariano. 1987. Cathodoluminescence and microprobe study ofrare earth elements in apatite. American Mineralogist 72:801–811.

Roscoe, S.M. 1969. Paper 68–40 in Huronian Rocks andUraniferous Conglomerates. Geological Survey of Canada.

Roskill Information Services Ltd. 1996. Rare Earths in the CIS.Report. London: Roskill.

Sadeghi, A., and K.F. Steele. 1989. Use of stream sedimentelemental enrichment factors in geochemical exploration forcarbonatite and uranium, Arkansas, U.S.A. Journal ofGeochemical Exploration 32:279–286.

Shepherd, M.S. 1990. Eneabba heavy mineral sand placers. Pages1591–1594 in Geology and Mineral Deposits of Australia andPapua New Guinea. Edited by F.E. Hughes. Parkville, Victoria:Australasian Institute of Mining and Metallurgy.

Sherer, R.L. 1990. Pajarito yttrium-zirconium deposit, OteroCounty, New Mexico. New Mexico Geology 12:21.

Smirnov, V.I. 1977. Ore Deposits of the USSR. Volume 3. SanFrancisco: Pitman.

Smith, M.P. 2001. The Bayan Obo Fe-REE-Nb deposit, InnerMongolia, China; comparisons with carbonatite-related and Fe-oxide-type deposits. Geological Society of America Abstractswith Programs 33(6):3.

Smith, M.P., and P. Henderson. 2000. Preliminary fluid inclusionconstraints on fluid evolution in the Bayan Obo Fe-REE-Nbdeposit, Inner Mongolia, China. Economic Geology 95:1371–1388.

Smith, M.P., P. Henderson, and L.S. Campbell. 2000. Fractionationof the REE during hydrothermal processes; constraints from theBayan Obo Fe-REE-Nb deposit, Inner Mongolia, China.Geochimica et Cosmochimica Acta 64(18):3141–3160.

Staatz, M.H. 1983. Geology and Description of Thorium and RareEarth Deposits in the Southern Bear Lodge Mountains,Northeastern Wyoming. Professional Paper 1049-D. Reston,VA: USGS.

Staatz, M.H., B.J. Sharp, and D.L. Hetland. 1979. Geology andThorium Resources of the Lemhi Pass Thorium District, Idahoand Montana. Professional Paper 1049-A. Reston, VA: USGS.


Steenfelt, A. 1991. High-technology metals in alkaline andcarbonatitic rocks in Greenland: Recognition and exploration.Journal of Geochemical Exploration 40:263–279.

Taylor, R.P., and P.J. Pollard. 1996. Rare earth elementmineralization in peralkaline systems; the T-zone-REE-Y-Bedeposit, Thor Lake, Northwest Territories, Canada. Pages 167–192 in Rare Earth Minerals: Chemistry, Origin, and OreDeposits. Edited by A.P. Jones, F. Wall, and C.T. Williams.Mineralogical Society Series 7. London: Chapman and Hall.

Taylor, S.R., and S.M. McClennan. 1985. The Continental Crust:Its Composition and Evolution. Oxford, UK: BlackwellScientific Publications.

Trueman, D.L., J.C. Pedersen, L. de St Jorre, and D.G.W. Smith.1988. The Thor Lake rare-metal deposits. Northwest Territories.Pages 280–290 in Recent Advances in the Geology of Granite-Related Mineral Deposits. Edited by R.P. Taylor and D.F.Strong. Montreal, Quebec: Canadian Institute of Mining andMetallurgy.

Vallee, M., and F. Dubuc. 1970. The St-Honore carbonatite deposit,Quebec. Transactions of the Canadian Institute of Mining andMetallurgy 73:346–356.

Van Wambeke, L. 1977. The Karonge rare earth deposits, Republicof Burundi: New mineralogical-geochemical data and origin ofthe mineralization. Mineralium Deposita 12:373–380.

Vlasov, K.A., M.V. Kuz’menko, and Y.M. Yes’kova. 1966. TheLovozero Alkali Massif. New York: Hafner.

Wakita, H., P. Rey, and R.A. Schmitt. 1971. Abundances of the 14rare earth elements and 12 other elements in Apollo 12 samples.Pages 1319–1329 in Proceedings of the Second Lunar ScienceConference. Edited by A.A. Levinson. Cambridge, MA: MITPress.

Wall, F., and A.N. Mariano. 1996. Rare earth minerals incarbonatites: A discussion centred on the Kangankundecarbonatite, Malawi. Pages 193–225 in Rare Earth Minerals:Chemistry, Origin, and Ore Deposits. Edited by A.P. Jones, F.Wall, and C.T. Williams. Mineralogical Society Series 7.London: Chapman and Hall.

Welch, B.K., J. Sofoulis, and A.C.F. Fitzgerald. 1975. Mineral sanddeposits of the Capel area, W.A. Pages 1070–1088 in EconomicGeology of Australia and Papua New Guinea. I: Metals.Monograph 5. Parkville, Victoria: Australasian Institute ofMining and Metallurgy.

Willis, M. 2002. A mountain of rare potential. Industrial Minerals415:63–65.

Woolley, A.R. 1987. Alkaline Rocks and Carbonatites of the World.Part 1: North and South America. Austin: University of TexasPress.

Wu, C., Z. Yuan, and G. Bai. 1996. Rare earth deposits in China.Pages 281–310 in Rare Earth Minerals: Chemistry, Origin, andOre Deposits. Edited by A.P. Jones, F. Wall, and C.T. Williams.Mineralogical Society Series 7. London: Chapman and Hall.

Yuan, Z., G. Bai, C. Wu, Z. Zhang, and X. Ye. 1992. Geologicalfeatures and genesis of the Bayan Obo REE ore deposit, InnerMongolia, China. Applied Geochemistry 7:429–442.

rare earth elements - sgtk - news · rare earth elements 771 table 2. estimated annual world mine...

Documents